1" 1‘11“» 111" ’11‘15-1' - """ ’ .~ 11111- " . "111111. 11 111 ‘5. ; .1 1111111 ‘ 3 “1‘11!” 11“ 1"11111‘111.ll 111111ng119: . ‘111'11‘111. 1‘11 111““1 1111111 “ 1‘11‘1 1,111,111: ”1'1‘1111‘1 1 1,11 1113111111 111 111‘ 1.131111'1" 1 1 1 1 ’ i}? I L‘ ' . . 2 1. . ' 119-: ' : . ‘ '1 " ‘1 .1“. 1 11 1%1a 11 1 1 K 11 1 11‘ I’ '1‘} 1113111: 1% 1%??? 11 .. '1 2;“ I, :1 1"}‘"'. "1 ' ..... -1‘Ji‘ '131111; ..... 5 . 1 l 1' 11 1 4 ,1 121. 1. t' . .; .. I," 1 1 5:: | r1111; & . 1,11 1 ,5 " 1:11;" "111111; 1:1 1" 51' 1 1'1 11.1 111111 " 11:11:11 - , :14 11:11)“ ' 1' '29. .1411 111111‘H'H ”111' ,1' . _ 3111:3411 1 1 :19? "-v. 111111 111.111.“ £11 1.1: "1111‘L11h‘ 1'11 1111111 ’"11'11171'1'1‘ '1 3, 11 11 m; 11111111111119 11112111111111.1111111111111-‘111 1111.1: ' ' 1 1. 1 ' 1iq'1‘111111115’1‘“r1111 {kuv1Jtl111w‘3-z1‘tr1‘m 9.411111 11 THESIS 1w.» .fi“. H“ Ha. .’ in» L 'h. l.- r This is to certify that the thesis entitled An Analysis of Psychotherapy Superv151on on the Dimen51ons of Complementarity, Process Interaction, and Perceived Relationship presented by Frank Marc Pasciuti has been accepted towards fulfillment of the requirements for Ph.D. (kg yeeui Counseling, Educa— tiona Psychology and Special Education (Major in Counseling Psychology) MW 4 Major professor Date 11 1981 0-7639 310382 5752 .,. _. . . IIIIIIIIIIZIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. t—JIHNZ‘TfiEEz’I'“ IL'NI 0 HAY o i 1.99? WW 1A% G‘A/Vu' AN ANALYSIS OF PSYCHOTHERAPY SUPERVISION ON THE DIMENSIONS OF COMPLEMENTARITY, PROCESS INTERACTION, AND PERCEIVED RELATIONSHIP BY Frank Marc Pasciuti A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Educational Psychology and Special Education 1981 ABSTRACT AN ANALYSIS OF PSYCHOTHERAPY SUPERVISION ON THE DIMENSIONS OF COMPLEMENTARITY, PROCESS INTERACTION, AND PERCEIVED RELATIONSHIP BY Frank Marc Pasciuti The primary purpose of this study was to investigate the communications and perceptions of the developing super— visory dyad. Examination was given to the dimensions of interpersonal complementarity in communications between supervisors and supervisees, the amount of therapeutic work interactions produced in those communications, and the per- ceptions of the facilitative conditions within the relation- ship. These dimensions were investigated independently for possible changes from the first to the second stage of super- vision and in relationship to each other within each stage of supervision. The stages of the developing supervisory relationship were arbitrarily defined as the terms in a three term academic year. The subjects in this study were ten doctoral practicum students and their supervisors. All subjects were affiliated with the Michigan State University counseling center training program. Two taped sessions were taken four times, at the beginning and the end of the first and second terms of the academic year. From the tapes, ratings were made on two Frank Marc Pasciuti measures. The first was a measure for behavioral comple- mentarity (Leary, 1957), and the second was a measure of the content and process levels of communication (Hill Inter- action Matrix, l96l). Lastly, a measure was taken on the perceptions of relationship in each stage by all participants (Barrett Relationship Inventory, 1959). The first hypothesis predicted a significant decrease in combined complementary interactions from Stage I to Stage II. The postulate was not supported by the data. Hypothesis II predicted a significant increase in therapeutic work re- sponses from Stage I to Stage II. Contrary results were found. In the third hypothesis, no significant changes were found to occur in combined perceptions of the supervisory relationship from the first to the second stage of super- vision. In the remaining three hypotheses, examination was given to the interrelationships among complementarity, thera- peutic work, and perceptions of relationship. The correla- tions were not significant in their predicted directions. The impact of high risk interactions upon perceptions of relationships and differences between therapy and supervision were discussed. The need to further clarify differences between therapy and supervision and monitoring supervision over a longer period of time were suggested for future re- search. DEDICATION To my Mother and Father. ii ACKNOWLE DGEMENTS I offer a deep appreciation to Dr. Douglas Miller. As chairman of my dissertation his direction, support and en- couragement kept the light at the end of the tunnel always in View. As a supervisor he provided a model of my ideal. But above all, his friendship was most cherished through my graduate career. I thank Dr. William Hinds for the direction he provided in establishing a blueprint for this study. And for his con- tinued guidance as I navigated the doctoral path. To Drs. Ralph Kron and Rebecca Henry go a warm thanks for their constant availability and enthusiastic support through the dissertation process. I thank Dr. Imogin Bowers, Sarah Woodhull, Dr. Joann Hamachek, Michelle Klee, Dr. Linda Forrest, Micky Green, Ted Bouterse, Bill Parker, Lew Dotterer, Paul Fatell, Pam Mont- gomery, Lila Sales, Linda Van Arsdale, Dr. Heide Joos, Jim Azar, Dave Rockwell, and Jerry Hermanson for contributing their time and a glimpse of their supervisory process to this study. Dr. Pricilla Hill, Becky Hollingsworth and Barbara Hol- lembeck for their help in completing the many tasks required in this study. iii Gabby for making the "Data-Day" struggles all "compute". A special thanks to Dr. John Powell for providing a compass when my ship was sailing stormy seas. Drs. William Mueller, Cecil Williams, Sam Plyler, Liz Monroe-Cooke and Mary Ann Stehr for their supervision and guidance, and for helping me better understand myself as a person and a professional. Russ and Bill for providing "retreats" through their friendship and love, and laughing away the tensions. Ron, Jim and Bob for the Opportunities to play and "sing" praise to the "higher" sides of life. Jane for her tireless editing and teaching, but most of all for her continued love. Finally to all at the Counseling Center, a special thanks for the stimulation and nurturance through my years at Michi- gan State. I continue to be with you all in spirit. iv TABLE OF CONTENTS Chapter Page I. LIST OF TABLES II. INTRODUCTION . . . . . . . . . . . . . . . . . 1 Problem . . . . . . . . . . . . . . . . . . 3 Theory . . . . . . . . . . . . . . . . . . . 7 Need . . . . . . . . . . . . . . . . . . . . 9 Purpose . . . . . . . . . . . . . . . . . . 11 III. REVIEW OF RELATED LITERATURE . . . . . . . . . 16 The Supervisory Process . . . . . . . . . . 16 The Relationship in Psychotherapy Supervision . . . . . . . . . . . . . . . l9 Behavioral Complementarity . . . . . . . . . 21 Complementarity in Therapeutic Relationships . . . . . . . . . . . . . . 26 Complementarity in Supervision . . . . . . . 29 Facilitative Conditions in the Supervisory Relationship . . . . . . . . 32 Facilitative Conditions in Super- vision and Trainee Self-Exploration . . . 34 The Supervisory Relationship as a Dyadal Unit . . . . . . . . . . . . . . . 37 Summary of Related Research . . . . . . . . 38 IV. METHODOLOGY . . . . . . . . . . . . . . . . . . 40 Sample Selection . . . . . . . . . . . . . . 40 Data Collection . . . . . . . . . . . . . . 42 The Barrett-Lennard Relationship Inventory . . . . . . . . . . . . . . . . 45 Behavior Analysis System . . . . . . . . . . 49 Interjudge Reliability During Training . . . . . . . . . . . . . . . 49 Categorizing and Scoring Verbal Interaction . . . . . . . . . . . . . . . Sl Rater Reliability on Hill Interaction Matrix . . . . . . . . . . 53 Hypotheses . . . . . . . . . . . . . . . . . 54 Analysis of Results . . . . . . . . . . . . 59 Results of Primary Concern . . . . . . . . . 59 V Chapter V. VI. VII. VIII. RESULTS 0 O O O O O O O O O O O O O Hypothesis I: Complementarity Hypothesis . . . . . . . . . Hypothesis II: Therapeutic Work Hypothesis III: Perceptions of Facilitative Conditions . . . Hypothesis IV: Relationship Be- tween Complementarity and Therapeutic Work . . . . . . Hypothesis V: Relationship Be— tween Complementarity and Super- visory Relationship Perceptions Hypothesis IV: Relationship Be- tween Facilitative Conditions and Therapeutic Work . . . . Summary of Results . . . . . . . DISCUSSION . . . . . . . . . . . . Subsidiary Discussion Section . Suggestions for Future Research Summary . . . . . . . . . . . . APPENDICES A. Barrett-Lennard Relation Inventory B. Scoring Manual for Interpersonal Behavior Rating System . . C. Hill Interaction Matrix Scoring Process 0 O O O O O O O I BIBLIOGRAPHY . . . . . . . . . . . Vi Page 61 61 64 67 70 77 84 91 93 116 121 123 127 130 136 142 Table 4.6 LIST OF TABLES The Interpersonal Circumplex . . . . . . . . Complementarity Matrix (Schiller Scoring System) . . . . . . . . . . . . . . . . . Interjudge Percentage of Agreement on Training Sample (Employing Behavioral Analysis System). . . . . . . . . . . . . Rating Reliability Percentages of Sampled Interactions (Employing Hill Inter— action Matrix). . . . . . . . . . . . . . Complementarity Index Scores for Each Supervisory Pair for Stage I (Fall Term) and Stage II (Winter Term). . . . . Complementarity Index Score Means, Change Scores and t—Values for Supervisory Pairs, Supervisors, and Supervisees Across Two Stages . . . . . . . . . . . . A Comparison of Hill Interaction Matrix (Quadrant Four) Mean Percentages and Difference Scores, for Individual Supervisory Pairs in Stages I and II . . A Comparison of Hill Interaction Matrix (Quadrant Four) Mean Percentages, Dif— ference Scores, and t-Values for Com- bined Supervisory Pairs, Supervisors, and Supervisees . . . . . . . . . . . . . Comparison of Combined Relationship Per- ceptions of Individual Supervisory Pairs (as measured by the Barrett-Lennard Relationship Inventory) for Stages I and II. . . . . . . . . . . . . . . . . . Comparison of Mean Relationship Perception (Measured by the Barrett-Lennard Rela- tionship Inventory), Difference Scores, and t-Values for Combined Supervisory Pairs, Supervisors, and Supervisees . . vii Page 23 27 51 54 62 64 66 67 69 7O Table 4.7 The Relationship Between Combined Levels of Complementarity andCom- bined Therapeutic Work Responses for Supervisory Pairs for Both Stages I and II and Illustration of an Approxi- mated Line of Best Fit for the Data Groupings . . . . . . . . . . . . . . . . The Relationship Between Bombined Levels of Complementarity and Com- bined Therapeutic Work Responses for Supervisors for Both Stages I and II and Illustration of an Approximated Line of Best Fit for The Data Groupings . The Relationship Between Combined Levels of Complementarity and Com- bined Therapeutic Work Responses for Supervisees for Both Stages I and II and Illustration of an Approximated Line of Best Fit for The Data Groupings Pearson Product-Moment Correlation Coeffi- cients for the Relationship Between Complementarity and Therapeutic Wrok Responses for Supervisory Pairs, Super- visors, and Supervisees During Each Stage Of Supervision . . . . . . . . . . The Relationship Between Cobined Levels of Complementarity and Perceptions of Supervisory Relationship for Super- visory Pairs for Both Stages I and II and Illustration fo an Approximated Line of Best Fit for the Data Groupings . . . . . The Relationship Between Combined Levels of Complementarity and Combined Perceptions of Supervisory Relationship forSuper- visors for Both Stages I and II and Illus- gration of an Approximated Line of Best Fit for the Data Groupings . . . . . . . The Relationship Between Combined Levels of Complementarity andCombined Perceptions of Supervisory Relatinship for Super— visees for Both Stages I and II and Illus- tration of an Approximated Line of Best Fit for the Data Groupings . . . . . . . viii Page 72 73 74 77 78 79 80 Pearson Product-Moment Correlation Coefficients for the Relationship Between Combined Levels of Comple- mentarity and Combined Perceptions of Supervisory Relationships Within Stages I and II forSupervisory Pairs, Supervisors, and Supervisees . . . . . . The Relationship Between Combined Thera- peutic Work Responses and Combined Perceptions of Supervisory Relationship for Supervisory Pairs for Both Stages I and II and Illustration of an Approxi- mated Line of Best Fit for the Data Groupings . . . . . . . . . . . . . . . . The Relationship Between Combined Thera- peutic Work Responses and Combined Perceptions of Supervisory Relationship for Supervisors for Both Stages I and II and Illustration fo an Approximated Line of Best Fit for the Data Groupings . . . The Relationship Between Combined Thera- peutic Work ReSponses and Combined Perceptions of Supervisory Relationship for Supervisees for Both Stages I and II and Illustration of an Approximated Line of Best Fit for the Data Groupings . Pearson Product-Moment Correlation Coeffi- cients for the Relationship Between Com- bined Percentage of Therapeutic Work Interactions and Combined Perceptions of Supervisory Relationship Within Stages I and II for Supervisory Pairs, Supervisors, and Supervisees . . . . . . . . . . . . . ix Page 84 86 87 88 91 INTRODUCTION The teaching and learning of psychotherapy through the supervisory process remains an integral part of the training of psychotherapists. Supervision is viewed as the prime component essential to training therapists across the many divergent schools of psychotherapy (Lambert, 1979). Regard— less of the various approaches to the process of supervision, it is the interpersonal relationship between supervisor and supervisee that serves as the primary channel through which learning occurs. Mueller and Kell (1972) note, "The relation— ship must develop between supervisor and therapist if the principal goal of supervision is to be achieved" (p.3). The importance of the interpersonal relationship in supervision parallels is significant to the relationship of client and therapist in therapy (Altucker, 1978; Rogers, 1957; Mueller and Kell, 1970). The development of an effective supervisory or therapeutic relationship is evidenced in its interpersonal communications (Dietzel, 1971; McElhose, 1973). These communications can indicate varying levels of anxiety and security maintenance. In relationships where risks are taken and anxiety is confronted rather than avoided, inter- actions are less complementary. 2 Complementary interactions serve to maintain security and minimize anxiety (Leary, 1957). The willingness to decrease the level of complementary interactions and move into more high risking exchanges may be supported by the facilitative conditions perceived in the relationship. Rogers (1957) views the development of any therapeutic relationship as depending upon the presence of these facili- tative conditions (empathy, level of regard, unconditionality of regard, and congruence). With these conditions perceived as being present by both parties, a foundation of trust may then develop and serve as a buffer to the sensitive work that must be interpersonally communicated (Berenson, 1967; Rogers, 1957; Walz, 1963). The interpersonal communications, facilitative condi- tions, and content of the work that transpires in supervision may vary in each relationship. The variance in these dimen- sions may be best understood when the relationship is examined as a dyadal unit. Lennard and Bernstein (1960) refer to the psychotherapy dyad as a "social system". Similar to the psychotherapy dyad, the supervisory dyad may offer insight into its process when investigated as a dyadic system. Within the supervisory dyad the process of supervision may change at different points in the progression of the developing relationship. Mueller and Kell (1972) suggest that supervision, like therapy, proqresses through three distinct stages of development. These stages are the follow- ing: the beginning stage where the process of supervision 3 is defined and trust development is a vital issue, the middle stage where much of the sensitive work that is critical to successful supervisory outcome takes place, and the ending stage when integration and termination are of primary concern. The present study investigates the levels of inter- personal complementarity in supervisory dyads over the first two stages of supervision. Further analysis is given to the facilitative conditions perceived in those relationships and to the amount of higher risking inter- actions produced. It is hOped that by examining these dimen- sions of the supervisory relationship greater understanding of the supervisory process may be achieved. This understanding may offer us insight into how supervisory dyads mature and master the sensitive and challenging work necessary for suc- cessful supervisory outcome. Problem Exploring sensitive issues in the supervisee's thera- peutic relationships, and within the supervisory relation- ship itself, will often be accompanied by elevated levels of anxiety. The ability to approach that anxiety and to utilize it constructively as an opportunity for learning may depend upon the level of facilitative conditions in the supervisory relationship. Mueller and Kell (1972) state, "To help a therapist to grow, the supervisor must often touch on the meaning of the therapist's behavior with his 4 clients; he must explore the anxiety he senses in the therapist; and at times he must awaken anxiety by pene- trating the rationalizations and defenses that bind it. The openness and collaboration necessary to permit growth can occur only if the supervisor and therapist trust each other" (p.30). The supervisee's movement towards growth may be side- tracked at times by conflicts arising in the supervisory relationship. These conflicts need to be confronted, and the accompanying anxiety needs to be approached rather than avoided. For an individual to utilize confrontations con- structively in a self exploratory manner, he must feel that the relationship has evolved to a point of security (Beren- son, Mitchell and Moravac, 1967; Anderson, 1968). Confrontation in the supervisory relationship may take the form of clarifying, pointing out strengths or weaknesses, and addressing discrepancies in perceptions. The necessity of confrontation by the supervisor is acknowledged by McElhose (1973), "The supervisor offers himself as a model of an effective and motivated therapist who is willing to confront the therapist-supervisee's anxiety and conflict with a genuineness and realism which permits the supervisee's self- exploration and experimentation" (p.12). Discrepancies that arise in the relationship of therapist and client, as well as supervisor and supervisee, need clarification, although addressing these conflicts may raise anxiety in both parties involved. 5 These confrontive interactions will not comprise the majority of the time spent communicating in the supervisory relationship, but their growth producing effect and the necessity of confrontation at appropriate times has been noted by many theorists (Ekstein and Wallerstein, 1959; Mueller and Kell, 1972; Does, 1969; McElhose, 1973; Ander- son, 1968; Berenson, 1967). Should the interactions between supervisor and supervisee move towards a more secure and less potentially stressful exchange, it is noted that the complementarity of those interactions may increase (McElhose, 1973; Dietzel, 1971). "Complementarity" refers to the amount of reward that is derived in interactions with participants. A highly complementary interaction would be one in which a great degree of reinforcement is achieved in response to another's elicitations. Sullivan sees complementary interactions as those where as individual strives for security and integra- tion. When encountering a person for the first time, and individual would be more likely to complement the other's interactions. These initial complementary interactions may benefit both parties by maintaining security and a low level of anxiety. Complementary interactions are non-confrontive and serve to minimize anxiety in relationships. Continual complementary interactions in therapy or supervision may indicate the avoidance of conflicts. Dietzel (1970) viewed therapy over three stages and found that those relationships where comple- 6 mentarity lowered during the middle stage (working stage) were positively correlated with success in therapy. It was inferred that as a result of reduced complementarity, security operations were lower, anxiety was more intense, and more therapeutic work was taking place. Similar to the therapeutic relationship, the supervisory relationship has been investigated in light of complementary interactions. McElhose (1973) came to the conclusion that, "The avoidance of dealing with anxiety could infer that a relatively high degree of complementarity will be present in the relation- ship. Should such an avoidance occur, the individuals involved will be reinforcing circularity of the conflict creating the anxiety, by eliciting from each other preferred response patterns, thereby preventing the potential threat of confrontations which could promote growth" (p.21). The present study addresses the question of why certain relationships evolve to a point where complementary inter- actions decrease and more high risk therapeutic work is undertaken. The underlying relational ingredients that serve to facilitate that movement have not been previously defined. More may be understood about this developmental process when investigated in light of the perceived facilita- tive conditions present in the supervisory relationship. In summary, the supervisory process is an important ingredient in the development of the psychotherapist. The interpersonal relationship between supervisor and supervisee is critical in serving as a foundation and model for learning. 7 The facilitative conditions within the supervisory rela- tionship might serve as a context within which risks can be taken and growth promoted. Movement in that direction may possibly be detected by lowering levels of complementarity in interactions, resulting in greater amounts of therapeutic work. Theorv The development of a sound supervisory relationship is similar to the development of any significant relation- ship in one's life (Mueller and Kell, 1971). The under- lying qualities that enhance and facilitate the establish- ment of trust are common to all. Rogers (1957) identified four crucial ingredients that are necessary in the develop- ment of any helping relationship. They are empathy, level of regard, unconditionality of regard, and congruence. The importance of these conditions in therapeutic and super- visory relationships, as well as within parent-child and teacher-student relationships, has been given considerable attention in recent years (Rogers, 1957; Barrett-Lennard, 1959; Anderson, 1967; Carkuff, 1969; Lambert, 1970; Pierce & Shauble, 1971; Karr and Gieist, 1977). In all relationships, the closer and more invested the involvement, the greater the probability of times of heightened anxiety. This potential in the supervisory relationship is acknowledged by Mueller and Kell (1971), "One of our premises about a supervisory relationship is 8 based in our more general observation that any signifi- cant human relationship is charged with anxiety" (p.8). The supervisory relationship has a diversity of potential for anxiety raising concerns. The trainee is invested in developing as a professional and will be evaluated in that regard. There will be times when personal issues relevant to clients will interact with similar personal issues in the trainee's emotional life. These situational and emo- tional anxiety fuelers will need to be addressed for the trainee to utilize them as opportunities for potential growth. The anxiety and stress that arise in addressing these sensitive areas will be experienced by supervisor and super- visee alike. At times the supervisee will reproduce in the relationship with his supervisor the same dynamics that his client is acting out in therapy (Mueller and Kell, 1971). Whether these anxiety producing dynamics are addressed or avoided may be detected in the interpersonal behaviors of the individuals involved. Leary (1957) states, "The inter- personal behavior of an individual is the machinery by means of which he wards off anxiety and maintains a multilevel balance of self enhancement" (p.15). Founded upon the inter- personal theories of Sullivan, Leary believes that truly understanding a person is to know the interpersonal techniques that he employs to avoid or minimize anxiety. Striving to achieve a better understanding of the super- visee, the supervisor may need to confront some of his 9 behaviors. These behaviors may be evidenced in the super- visee's interactions with his clients or in their super- visory relationship. In these confrontive situations the supervisory process may closely resemble therapy, requiring greater personal risk and investment. The timing and delivery of such confrontations will be crucial to their effective utilization (Ekstein & Wallerstein, 1958). The self exploration that may follow these confrontations will be dependent upon the facilitative conditions present in the relationship (Anderson, 1967). Confrontations of such a personal nature will not occur in supervision as frequently as they do in therapy (Lambert, 1971). However, the freedom to express these confrontations and their constructive utilization may be enhanced or lessened due to the conditions offered in the supervisory relationship. Need Although many professionals involved in the training of psychotherapists have acknowledged the importance of the supervisory process, little research has been directed towards the interpersonal relationship between supervisor and supervisee. Further, less attention has been focused on their interactions or the ingredients within the relation- ship that might facilitate growth in the trainee. Re- searching interview behavior and relationships in counseling supervision, Santoro (1970) note, "It must be remembered 10 that the supervisory relationship itself is a human inter— action process and as such the dynamics of supervisor and supervisee are reciprocally affecting and changing each other. If the relationship is important to successful outcomes, it is necessary to examine carefully the condi- tions under which effective relationships are achieved in the process" (p.11). When investigating the components of effective super- visory relationships, Hansen and Warner (1971) pointed to the lack of identified supervisor behaviors and their rela- tionship to trainee behaviors. Wedeking and Scott, (1976) following an investigation of supervisory relationships on the dimensions of empathy and other specific trainee competencies, suggested that further investigations into the process of supervision, the systematic descriptions of the teaching and learning processes and students involved, are warranted. Despite the fact that the relationship between super- visor and supervisee is acknowledged as an important ingre- dient to the supervisee's professional development, little attention has been directed towards the behaviors that facilitate that process. Hansen and Barker (1962) investi- gated the extent to which the level of supervisor-trainee relationship was related to the trainee's level of experi— encing. They inferred that trainees who viewed their super- visors as congruent, having a high level of regard, empathetic understanding and unconditional positive regard 11 felt freer to examine their feelings and experiences. They suggested future research "to spedify further the separate types of supervisor behaviors and evaluate their relevance to counselor education" (p.109). The ingredients of effective supervisory relation- ships need further investigation. The behaviors of both supervisor and supervisee, as well as the inspection of their interpersonal communications, must be identified. The components of a good working relationship and their impact on the enhancement of supervisee growth must also be monitored. Purpose It is the purpose of this study to investigate the factors that may influence the progression of the super- visory relationship from its early developing stage to its middle working stage. The investigation will be focused on the measurement and analysis of the interrelationships among: (1) the interpersonal level of complementarity in ongoing supervisory relationships from their early develop- ment through the middle (working) stages; (2) the perceived quality of those relationships on the dimensions of empathy, level of regard, congruence, and unconditional positive regard as reported by both participants; and (3) the amount of confrontation (therapeutic work) that is produced by each supervisory pair. 12 In past studies of supervisory relationships, findings have confirmed that the developmental progression occurs over a fairly definable sequence of stages (Kell & Mueller, 1972; McElhose, 1973). Similar to therapeutic relation- ships, changes have been noted in the interpersonal communi- cations that occur between supervisors and supervisees over the course of these stages. The early stages of any rela- tionship are accompanied by high levels of complementary interactions as security operations are maintained and trust is initially developing. Mueller (1969) and Dietzel (1971) found that in the late stage of therapy, a time of integra- tion and termination, complementary interactions are also at high levels. However, in the middle stage of therapy it was found that lower levels of complementarity significantly distinguished (p(.005) successful therapy from unsuccessful therapy outcome (Dietzel, 1971). Behavioral complementarity according to Leary (1957) and Carson (1969) refers to the degree of reward (i.e. inter- personal reinforcement) experienced by both interaction participants as a result of the particular behaviors ex- changed. The Leary Interpersonal Circumplex (see Appendix A) is used as a basis to judge the degree of complementarity in the supervisor-supervisee relationship in this study, expressed as a CI (Complementarity Index). According to this Circumplex, interpersonal interactions are reciprocated along dimensions of a dominance-submission axis and a friendly- hostile axis. Complementary interactions serve to maintain 13 security and minimize anxiety. The level of security in any interaction is influenced by the degree of trust one individual feels towards another. Trust as a basic component in any relationship is of prime importance in the therapeutic and supervisory relationship if risks are to be taken. The trust necessary for taking therapeutic risks is dependent upon certain facilitative conditions present in the helping relationship (ROgers, 1957; Walz, 1963; Berenson, 1967). But the presence of these conditions is not enough; the conditions must be perceived as being present by the indivi- duals involved (Kurtz and Grummon, 1970; Rogers, Gendlin and Truax, 1967). An interest of this study is the per- ceived quality of the supervisory relationship as it relates to the amount of therapeutic work that transpires. Therapeutic work in this study refers to those higher risk interactions where supervisor and supervisee are focusing on their relationship or where the supervisee is confronted about his personal issues as they relate to both his client and the supervisor. These interactions, due to their highly personal nature, are often accompanied by higher levels of anxiety but hold a potential for greater personal awareness and growth for the professional. In order for these confrontations to be assimilated and used construc- tively, certain conditions need to be present in the relation— ship. Barrett-Lennard (1959) deve10ped an instrument that measures these "necessary and sufficient conditions for therapeutic change". The presence of these conditions in the l4 supervisory relationship have been proven to have a positive relationship to the trainee's level of experiencing (Hanson and Barker, 1963). This study makes use of the Barrett- Lennard Relationship Inventory (RI) to measure perceptions of both supervisor and supervisee. As a final area of investigation, this study focuses on the "product" of the interactions between supervisor and supervisee. Although the amount of therapeutic work tran- spiring in supervision may be much less than that taking place in therapy (Lambert, 1974), the confrontations and self-explorations that comprise this area are noted as necessary for satisfactory supervisory outcome (Does, 1968; McElhose, 1973; Mueller and Kell, 1972). Confrontive interactions between supervisor and super- visee will not only raise anxiety and stress levels but offer a growth producing element in the learning experience. Kell and Mueller (1966) state, "A supervisory relationship that is free of stress is one which is contemptuous of the coun- selor's strength and does not extend the counselor so that he will grow maximally" (p.108). The use of confrontation is an essential tool in the supervisory relationship, but it must be timed correctly and done within the establishment of a good relationship (Ekstein and Wallerstein, 1958). In a study of the comparison of supervision and coun- seling, Lambert (1974) made use of the Hill Interaction Matrix (HIM) deve10ped by Copplino and Hill (1959) for scoring interactions in those relationships. The HIM is used to 15 conceptualize interactions among two dimensions, content (what is being discussed) and process (how it is being discussed). The HIM is used in this study with specific interest in those interactions that fall into four areas on the matrix. These are (1) personal confrontive, (2) relationship confrontive, (3) relationship speculative, (4) personal speculative. These four areas of interaction are noted by Hill to assume "therapeutic work" is taking place. In summary, the purpose of this study is to investigate the interpersonal interactions in the supervisory relation- ship as they relate to the level of complementarity and the amount of therapeutic work taking place. Differences are expected to be found in these two dimensions as they occur in the context of the perceived quality of the supervisory relationship reported by both participants. CHAPTER II REVIEW OF RELATED LITERATURE In recent years researchers have attempted to fill the void of research on psychotherapy supervision. Much of the attention given to this undertaking has been directed towards investigating the variance between the didactic versus the experiential approaches. Advocates of the experiential approach closely parallel supervision to the therapy process itself. Greater emphasis and concern is placed upon the nature of the supervisory relationship. This chapter focuses on those studies dealing with the supervisory relationship, particularly upon the examination of interpersonal communica- tion, facilitative conditions offered, and the content of communication. More detailed discussion will be given to research most closely related to this author's study. The Supervisorngrocess Learning to understand the behaviors of others and of oneself in the psychotherapy process is a task of paramount importance. The therapist in training will often need to depend a great deal on the quality and productiveness of the training Opportunities available to him. Psychotherapy 16 l7 supervision is the vehicle through which that learning can occur, and it may be available in various forms, individual or group. A primary goal in supervision is expanding the trainee's awareness of the various aspects in the psychotherapy pro- cess. The ability to differentiate the clients' dynamics and behaviors and their interaction with the therapist's reSponses is a skill continuously being refined. In addi- tion to specifically expanding and differentiating the train— ee's awareness of the psychotherapy process, supervision more broadly encompasses such activities as consultation, coun- seling, training and instruction, and evaluation (Boyd,l978). Some of these supervision activities are given greater em- phasis depending upon the particular supervisory approach one endorses. The two most common streams of thinking dichotomize the supervisory process into two models, the didactic and the experiential. The didactic and experiential approaches have gained considerable attention in the research that has been done on supervision. Such involvements as providing information, directing, giving advice, and evaluating are generally con— sidered to be more didactic in nature (Krasner, 1962; Krum- boltz, 1976; Wolberg, 1954). These activities are more generally considered "teaching" within the supervisory rela— tionship. In contrast to this didactic style of supervision, supporters of the experiential approach contend that the supervisor should facilitate supervisee self exploration of 18 concerns and feelings both in regard to the therapy rela- tionship and in their supervision relationship (Rogers, 1957; Mueller and Kell, 1970; Patterson, 1964; Altuker, 1967). These advocates maintain that the didactic approach exercised alone becomes narrow and mechanistic. The re- buttal by the didactic supporters contends that the experi- ential approach subjects the supervisee to a role of client, creating unnecessary dependancy and undermining the collegial status in the relationship. The research undertaken to identify the empirical sup- port for one approach over the other has been very conflicting and inconclusive. Many researchers think it might be best to integrate both facets of supervision, implementing the parti- cular activities prescribed when deemed appropriate (Truax and Charchuff, 1967; Lister, 1963). Trembly (1981) suggests that emphasis upon the didactic approach may be more appro- priate at early stages of therapist training, while experien- tial involvement may be necessary at later stages of develop- ment. The supervision process entails many activities very important to the growth of the supervisee. Regardless of which approach a supervisor may think best serves the training needs, there is no clear support for one approach over the other. Arbuckle (1966), and Patterson (1964), believe that an attempt to integrate both approaches may be the most fruitful. Where the didactic approach can offer the teaching of skills, the experiential can facilitate trainee self l9 awareness. In the latter approach, similar to therapy, greater emphasis is placed upon the quality and conditions surrounding the supervision relationship. The Relationship in Psychotherapy Supervision Within the experiential component of supervision, the relationship between the supervisor and supervisee is the medium through which the learning occurs (Ekstein and Waller- stein, 1958; Mueller and Kell, 1972). By acknowledging the utility of the relationship, consideration may be given to the conditions within the relationship, interpersonal styles, and the content of communications. The relationship factor in supervision then warrants the same consideration as the relationship factor in psychotherapy. Supervision may be viewed as "therapeutic" even though much of the learning that occurs may involve cognitive elements (Boyd, 1978). Super- vision is basically a human relationship; it will have its moments of therapeutic exchange, as well as the anxieties that frequently accompany close interpersonal relationships (Patterson, 1964; Mueller and Kell, 1972). Describing the supervisory relationship Margolis (1970) states, "For me, the Optimal supervisor-supervisee relationship is no different than that which occurs between the therapist-patient at the times when growth is allowed to occur" (p.54). Operating with this analogy Of supervision and therapy in mind, the supervisor may then attend to the relationship as an important component in the learning process. 20 Frequently, the supervisee will interact with the super- visor in the same way that his client interacts with him in therapy (Mueller & Kell, 1973). The supervisee will replicate right within the supervisory relationship the same dynamics that is of central importance to his client. McElhose (1973) mentions: "Through the supervisor's aware- ness of how his supervisee is interacting with him, valuable information can be gained concerning supervisee dynamics and how these dynamics affect his ability to promote client change" (p.8). When these interactions between the supervisor and supervisee reflect what is transpiring in the trainee's relationship with his client, the supervisor needs to focus on their own relationship for learning to occur. At other times the supervisor may gain awareness into the supervisee's personal dynamics evidenced in his interaction with his clients. Addressing these concerns and other possible countertransference issues will often require confrontation within their own relationship. In all of these situations the supervisor and supervisee may experience greater anxiety within their relationship. The management Of anxiety within the supervisory rela- tionship will not only influence the effectiveness of super- vision but serve as a model for the supervisee in his therapy. The significance of appropriate handling of anxiety in supervision is commented upon by Mueller and Kell (1972); "whetheraatrainee therapist approaches anxiety in himself 21 and others, whether he aVoids anxiety in himself and others, or whether he binds anxiety in himself and others are most significant dimensions in determining the possible level of development of a therapist" (p.204). The trainee's capa- bility to manage anxiety has implications for his competency in doing therapy. Therefore, the function of supervision assumes the supervisor's responsibility to demonstrate productive methods of managing anxiety within their rela- tionship. Anxiety in the supervisory relationship as in all rela— tionships is evidenced in the interpersonal communication between the participants (Sullivan 1953). When this anxiety is avoided and security is maintained, the interactions between participants become much more complementary (Leary, 1957). Behavioral Complementarity Complementarity in interpersonal interactions refers to the amount of reward that is derived by both participants in the behaviors exchanged (Dietzel, 1971). Complementary interactions serve to minimize anxiety and maintain security. When the interactions between participants are not rewarding to one or both individuals they will be depicted by lower levels of complementarity. The concept of complementary interaction has its roots in the Sullivanian school of interpersonal theory (Sullivan, 1953). The individual is perceived as interacting with the environment from the earliest days of infancy. Shiller (1977) 22 elaborates upon this point stating, "The infant's cry has a definite impact on those within hearing distance. He has the ability to arouse some level of anxiety in those who are responsible for him, and they will respond with behavior that will effectively reduce that anxiety. The cry and re- sponse are the prototypes in interpersonal behavior" (p.5). As the child matures he becomes aware that exhibiting certain behaviors brings about particular responses from others. These eliciting behaviors and responses become an integrated part of the child's interpersonal repertoire. The accompany- ing nonverbal and affective components of these behaviors add to the total makeup of interactional patterns and strengthen the power of eliciting the desired interpersonal responses. Freedman, Leary, Ossorio and Coffey (1951) described this interpersonal function, calling the particular behaviors "interpersonal mechanisms". An individual learns to employ a number of interpersonal mechanisms. These mechanisms be- come more familiar with repeated usage, and they can be depended upon to pull or provoke reciprocal behaviors from others. These reciprocal responses are called interpersonal reflexes (Leary, 1957). 23 TABLE 2.1 DOMINANT Friendly Dominant ‘Quadrant Hostile, Dominant Quadrant Punish ). 1‘3 a’ - Z 5 ‘1‘ a as Cooperate F Hostile I. G ’ Condemn Friendly Submrssnve Self Submissive - Quadrant Quadrant SUBMISSIVE THE INTERPERSONAL CIRCUMFLEX 24 Reviewing the systems for analyzing and classifying these interpersonal behaviors Carson(l969) concluded: "On the whole, the conclusion seems justified that major portions Of the domain on inter- personal behavior can profitably and reason- ably accurately be conceived as involving variations on two independent bipolar dimen- sions. One of these may be called a dominance- submission dimension; it includes dominant, assertive, ascendant, leading, controlling (etc.) behaviors on the one hand, and submissive, re- tiring, obsequious, unassertive, following (etc.) behaviors on the other. The poles of the second principle dimension are perhaps best approximated by the terms hate versus love: The former in- cludes hateful, rejecting, punishing, attacking, disaffiliative (etc.) behaviors, while the latter includes accepting, loving, affectionate, affilia- tive, friendly (etc.) social actions" (p.102). Organizing these interpersonal behaviors into a rating system, Leary (1957) and his research associates at the Kaiser Foundation deve10ped an interpersonal Circumplex. In this system behavioral complementarity occurs on the basis of reciprocity in respect to dominance-submission and on the basis of correspondence in respect to hostility- affection. These dimensions serve as bipolar axes in the Circumplex around which certain behaviors are symmetrically categorized. Formulating his theory of the "principal of interper- sonal relations", Leary discusses interactions as inter- personal reflexes which "tend (with a probability signifi- cantly greater than chance) to initiate or invite reciprocal interpersonal responses from the other person in the inter- action that led to a repetition of the original reflex" (Leary, 1955, p.148). To illustrate the principal of 25 complementarity, one may assume an interaction between a supervisor and supervisee as rated on the Leary Circumplex (detailed discussion can be found in Appendix A). If the supervisor emits a friendly-dominant (FD) response (e.g. the supervisor asks the supervisee how his week developed with clients), and the supervisee responds with a friendly— submissive (FS) response (he cooperates responding with a statement that his therapy went well and then elaborates). The interaction is conceived as mutually rewarding and therefore highly complementary. If following this friendly- dominant response by the supervisor the supervisee responded with a hostile-submissive (HS) response (complains or condemns himself), or a Hostile-dominant (HD) response (re— jecting or boasting), the degree Of inferred reward (comple- mentarity) is less. The anxiety in the latter situation would be higher, therefore disturbing the maintenance of security systems. Research reviewed later in this chapter will indicate that at certain times it is necessary to move into less complementary interactions. In order to evaluate behavioral complementarity as a process, Dietzel (1971) developed and operationalized a quantitative measure categorizing the weightings for all the possible interaction combinations. This "interaction cell" was further refined by Schiller (1977), and implemented in the scoring of complementarity in this study. A 4 x 4 matrix was constructed with each of the 16 cells assigned weightings to reflect the level of complementarity. 26 Complementarity in Therapeutic Relationships Monitoring the changes in complementarity within a therapeutic relationship may offer the researcher insight into the ebb and flow of anxiety maintenance and security Operations. Dietzel (1971) investigated the varying levels of complementarity in therapeutic relationships over the course of therapy. It was hypothesized that as these rela- tionships matured and the participants engaged in more anxious periods of therapeutic involvement, their levels of complementarity would decrease. The complementarity index was employed to rate the interactions, and predictions were made to identify successful therapeutic outcome. Success in therapy was measured via pre-to-post-therapy MMPI pro- files for 20 psychotherapy dyads divided into ten successful and ten unsuccessful outcome groups. The interactions occurring in these therapy dyads were sampled from early, middle, and late stages in therapy. The middle stage of therapy was of particular interesttxJDietzel's study. Identified as the "working stage" by Mueller and Kell (1970), Dietzel expected that lower levels of complementarity would accompany successful outcome groups at that middle stage. Significant results were found in predicting successful from unsuccessful therapy groups (p(.005). Dietzel inferred that as therapeutic work transpired in the middle phase of therapy, anxiety increased as indicated by lowering levels of comple- mentarity. The explanation for the higher levels of comple- mentarity in the unsuccessful therapy dyads during the middle a b 27 TABIEIZ. 2 Complementarity Matrix (Schiller Scoring System)8 Hostile Dominant Friendly Dominant Friendly Submissive Ellcltor Behavior ' Hostile Submissive Respondent Behavior Hostile Friendly Friendly Hostile Dominant Dominant Submissive Submissive (leb (up um um: (l)p (2)9 (4)p (3)9 (3)9 (h)p (2)9 (llp (hlp (3)p (llp (2)9 Complementarity score =2 (Ecolumns). (X) a cell weight; p = proportion of cell interactions. 27 TEUKLE 2. 2 Complementarity Matrix (Schiller Scoring System)3 Respondent Behavior Hostile Friendly Friendly Hostile Dominant Dominant Submissive Submissive Hostile b Dominant“ (2)9 (l)P (3)9 (4)p I— o '; Friendly .2 Dominant (”P (2)9 (‘09 (3)p 8 3 . u Friendly :2 Submissive (DP (MP (2)P (”P G ‘ Hostile Submissive (h)p (3)9, (i)P (2)p bComplementarity score ‘ = 2 (Ecol umns) . (X) a cell weight; p - proportion of cell interactions. 28 phase of therapy was interpreted as the therapist getting "caught up" into the earlier behavioral elicitations, resulting in a maintenance Of less threatening interactions. Investigating changes in therapeutic relationships over time is necessary for a true picture of interpersonal move- ment. Lennard and Bernstein (1960) describe the psycho- therapy dyad as a social system with a tendency towards "homeostasis", minimizing levels of interpersonal stress. They assert the importance of measuring interpersonal beha- viors at various points in time, avoiding the possibility of only sampling an incidental time Of homeostatic balance. If for example one measure of behavioral complementarity was derived at a point in time, it would not be truly indicative of the changes that may be occurring in the ongoing process. Mueller (1969) compared interpersonal responses of clients and therapists at early and later phases of therapy. Employing the Leary system, he grouped portions of behaviors for each therapeutic dyad into quadrants during each phase of treatment. His results supported the notion that comple- mentarity generally increased at initial and late stages of therapy. Other implementations of the theory of interpersonal complementarity have been applied to therapeutic relation- ships. Researchers such as Raush, Dittman, and Taylor (1969), Mackenzie (1968), Crowder (1970), Mueller and Dilling (1969), and Schiller (1977) have found the Leary model sensitive to interpersonal communications in therapeutic relationships. 29 The research empirically supports expected changes in com- plementarity at different intervals during the therapy process. In order for relationships to grow both in therapy and supervision, it seems that a reduction in complementarity must occur at apprOpriate times to stimulate approaching anxiety and further facilitate a differentiation process between participants (Dietzel, 1971; McElhose, 1973). Complementarity in Supervision Despite the many applications of the concept of inter- personal complementarity in measuring therapeutic relation- ships, little exploration has been undertaken to specifically investigate the supervisory relationship in its interpersonal communications. McElhose (1973), however, did utilize the Leary system in his investigation of the supervisory rela- tionship. Monitoring groups of six experienced and six in- experienced supervisors on their levels of complementarity over a three term sequence of psychotherapy supervision, McElhose expected that noticeable differences would occur over time and between levels of experienced supervisory groups. A measure of weighted relational distance (WRD), the interactive distance or closeness between participants, was employed to further explain changes in complementarity. Though no significant differences were obtained between ex- perienced and inexperienced groups on the Complementarity Index (CI) or the WRD measure, there was a trend for inex- perienced supervisors to exhibit higher mean CI scores. 30 McElhose concluded that experience was not a good indicator for determining changes in complementarity over time. In a second facet of his study, McElhose expected that levels of complementarity between participants would de- crease in the middle phase of supervision. Similar to the middle "working" stage of therapy, this phase in the super— visory process was expected to accompany more anxiety and confrontation than the early and later stages. Further, the assumption was made that experienced supervisors would be less complementary during the middle phase, and more likely to engage in greater amounts of anxiety producing and confrontive interactions than their less experienced counterparts. McElhose (1973) maintained that in order for these confrontive and anxious interactions to occur, comple- mentary exchanges would have to decrease. The results of his research indicated that there were no significant dif- ferences based on levels of experience, but a trend toward non-complementarity by the experienced supervisors was apparent. The possible inverse relationship between levels Of interpersonal complementarity and more anxious confrontive interactions is of primary interest to this author's study. Although actual inspection of the content of interactions was not undertaken in the Dietzel (1971) study, a major inference was made that "therapeutic work" interactions increased when complementarity decreased. However, super- vision may not encompass the degree of "therapeutic work" that therapy itself entails. 31 Lambert (1974) compared the supervision and counseling processes, and his results supported this notion. He examined the behaviors of five professional therapists per- forming both supervision and therapy. Their levels Of facilitative conditions Offered (empathy, respect, genuine- ness, and specificity) and amount of therapeutic response statements spoken (measured by the Hill Interaction Matrix) were measured in the two contexts and then compared. Fol— lowing statistical analysis, supervision proved to be signi— ficantly less therapeutic than therapy. Twice as many state- ments were classified as having high therapeutic value during counseling than during supervision. In the comparison of facilitative conditions Offered in both situations, level Of respect and genuineness were equally evident, while empathy was not as apparent in supervision as in therapy. Lambert pointed out that although supervisors may have taught their supervisees to be empathic with their clients, supervisors were less empathic to their supervisees by comparison to their own clients. It might be inferred from this study that althoughtflmnxawere fewer therapeutic statements made in super- vision when compared to therapy, other variables, such as respect, genuineness and specificity, may set the stage for such encounters. These conditions may have equal importance to both therapy and supervision. 32 Facilitative Conditions in the Supervisory Relationship Providing facilitative conditions in both therapeutic and supervisory relationships has been related to the therapeutic effectiveness of those relationships (Rogers, 1957; Barrett-Lennard, 1959; Anderson, 1968; Berenson, Mitchell, and Moravec, 1968). Rogers (1957) professed that conditions such as empathy, genuineness, positive regard, and unconditionality Of regard are necessary and sufficient to any therapeutic relationship. Investigating therapist response as a causal factor in therapeutic change, Barrett-Lennard (1959) developed an instrument (The Barrett-Lennard Relationship Inventory) to measure the "necessary and sufficient conditions" for thera- peutic change postulated by Rogers (1957). The instrument provided a score for measuring a person's perception of another individual's interactions in respect to the condi- tions incorporated in the inventory. In this study comparing client perceptions and change in therapy, clients were asked to respond to the questionnaire on how they perceived these facilitative qualiteis in their respective therapists. A significant relationship was indicated between the perceived degree of facilitative conditions and the amount of change incurred by the client. The presence of these facilitative conditions in the supervisory relationships may not only enhance the relation- ship, but when exhibited by the supervisor, provide a model for their development in the trainee. The transferential 33 effect of supervisor-offered conditions in supervision and their adOption and application by trainees in therapy was examined by Pierce, Carchuff, and Berenson (1967). Their study indicated that the supervisors who offered highest levels of facilitative conditions elicited the greatest amount of constructive gain in trainees' demonstration of the same facilitative conditions. The effect of supervisor-offered conditions on the develOpment of supervisee functioning in therapy is further supported by Karr and Geist (1977). Studying newly—formed supervisory dyads over a 15 week period, supervision sessions were tape-recorded at the first, eighth, and fifteenth week intervals. The Barrett-Lennard relationship inventory was completed by the trainees immediately following the first and fifteenth supervision sessions. The trainees also sub- mitted recordings of their following therapy sessions. The supervision and therapy sessions were rated by judges on scales deve10ped by Carcuff and Berenson (1967) measuring variables of empathy, genuineness, specificity, and regard. Results indicated that genuineness, specificity, and respect offered by supervisors in supervision were positively related to how trainees manifested these conditions in their therapy. On the other hand, trainee "perceptions" of supervisor- Offered conditions in supervision had no relationship to the degree of facilitative conditions offered in trainee therapy. Pierce and Schauble (1970) also examined the impact of supervisor functioning upon intern trainees. The facilitative 34 conditions demonstrated by supervisors in their therapy relationships were measured. The results of Pierce and Schauble's study revealed that interns changed considerably and positively when supervised by therapists functioning at high levels of facilitative conditions, while interns who were supervised by lower functioning supervisors did not change or decreased slightly. One criticism Of this study, however, is that the supervisor conditions were measured while they were functioning as therapists. Lam- bert (1974) showed how these conditions may vary in the same person when he is engaged in therapy as Opposed to supervision. The studies cited above indicate that the facilita- tive conditions offered by supervisors in supervision relate to trainee adoption, development, and implementation of the same conditions in their therapy. Not only is there a "carryover" effect by trainees into their therapy, but the presence of these conditions in supervision may further en- hance the supervisory relationship establishing a conducive environment for other areas of trainee growth. Facilitative Conditions in Supervision and Trainee Self-ExploratiOn The necessary presence of facilitative conditions and their influence in the therapy process may be equally impor- tant to the supervisory process if it gravitates towards a more experiential or "therapy like" mode (Gray, 1974). It seems that when the supervisory process necessitates 35 addressing such issues as trainee countertransference reactions to client of supervisor, differences between supervision and therapy are less clearly differentiated. Perhaps ijzit; an even riskier situation for a supervisee to explore personal areas in supervision than for a client to do so in therapy. The client is there for that specific purpose, whereas the trainee is involved in a process that encompasses a broader spectrum of activity. In their research on student reactions to supervision, Miller and Getting (1966) surveyed trainee perceptions of supervision and found that the student feels at the mercy of the supervisor. The student needs to believe that the supervisor values him as a person as well as a counselor. Students sometimes resent the supervisor who responds to them strictly as a therapist. However, sometimes a super- visor may need to venture into the trainee's personal con- cerns that may be inhibiting the effectiveness of his therapy (Mueller & Kell, 1972). Does (1969) investigated psychotherapy supervision on dimensions of confrontation and levels of supervisor Offered conditions. A significant result of his study was that supervisors who rated high on empathy, positive regard, genuineness and concreteness experientially confronted their trainees more frequently than low rated supervisors. This result was similar to that of Anderson's (1968) and Beren- son's et al (1968) studies where it was found that therapists who scored high on facilitative conditions offered, 36 confronted their clients more frequently than therapists who scored low. These latter studies also revealed that due to the presence of these conditions, clients tended to further self explore following confrontations by thera— pists high on conditions offered. The further self exploration included personal material that had previously been unrevealed. This possible willingness to further self explore and utilize confrontations when initiated by a supervisor high on facilitative conditions may be evident in the supervisory relationship. Commenting upon conditions surrounding the supervisory process, Gray (1974) states, "the supervisor needs foremost to be supportive of the student counselor in his process of growing. He needs to communicate empathic understanding of the student's vulnerability and struggle, and needs to let him know that he is valued personally. At the same time, however, the supervisor needs to be honest about the student's counseling performance and how he might be affecting clients" (p.13). These confrontations may be anxiety producing for both supervisor and trainee but, as Gray points out, within a caring relationship the threat can be minimized and a balance between support and confrontation can be achieved. The alternative to confronting the trainee might be passivity and avoidance of such anxiety provoking encounters. Mueller (1972) feels that this would be "contemptuous of the trainee's strength", because by confronting the super- visee when deemed appropriate we are conveying a sense of 37 respect and caring. Rice (1974) expresses this message to the supervisee in his way by stating, "I know that you are strong enough to absorb my criticism and are capable of behaving better as both a counselor and a person" (p.19). The presence of facilitative conditions may enhance a supervisory relationship as they increase the effective- of a therapeutic relationship. These conditions when offered by a supervisor can serve as a buffer to the sometimes sensitive work that will transpire in supervision. Though a supervisory relationship will not always be engaging in confrontation and evaluation, these activities do occur in the learning process and may be best received when mediated within a caring supportive relationship. The Supervisory Relationship as a Dyadal Unit Investigating the supervisory relationship in light of its interpersonal communications may necessitate approaching the relationship as one interdependent unit. The elicita- tions and responses that accompany interpersonal communica- tions are largely the results of patterns learned in response to another person (Sullivan, 1953). One person's response to one individual may be very different than that same per- son's response to another individual. When examining complementarity in interpersonal communications the researcher analyzes one person's response to the preceding person's elicitation. This process emphasizes the interdependent connectedness of interpersonal communications. Dietzel (1971) 38 used the therapy dyad as his unit of analysis when investi- gating complementary interactions and their relationship to therapeutic outcome. Related evidence of the interdependent qualities of communications are provided by those researchers comparing facilitative conditions offered by one individual and the amount of interpersonal risk taken by another (Anderson, 1968; Berenson, Mitchell, and Moravec, 1968). Lennard and Bernstein (1970) describe the psychotherapy dyad as a "social system". Equally analogous to the supervisory relationship Kell and Mueller (1972) state, "The dynamics of the super- visee and supervisor are reciprocally affecting and changing each other" (p.129). They go on to say that the deepening of the relationship leads to an "interlocking" Of expectations. In the present study, approaching the supervisory rela- tionship as a dyadal unit is Of primary importance. The complementarity Of interpersonal communications is dependent upon two individuals interacting. The perceived quality Of facilitative conditions and the inclination to venture into higher risking interactions will also be affected by what both individuals bring together in their relationship. Summary The research reviewed in this chapter indicates that there is no clear evidence for the exclusive adoption of one supervisory mode. The supervisory process may be best served when employing a combination Of both experiential 39 and didactic approaches. When the process includes more experiential activities, the relationship becomes the medium of their expression. The interpersonal communica- tion between participants can give indication of anxiety maintenance and security Operations by their levels of interpersonal complementarity. Complementary interactions will be high at the begin- ning and ending of therapeutic and supervisory relationships, but if the work of both processes is to be achieved, the level of complementarity will need to decrease at working stages. This "therapeutic like" work in supervision may take the form of confrontation and personal self exploration. As in the therapy process, such interactions have been found to be best mediated and more likely to occur when accompanied by high levels of facilitative conditions. These facilita- tive conditions when Offered by a supervisor will not only serve to model and teach the trainee how to adOpt them but will also enhance the supervisory relationship. The investi- gations of all those variables under consideration may best be accomplished when viewing the supervisory relationship as a dyadal unit with both individuals contributing to the same social system. CHAPTER III METHODOLOGY In the present study differences are examined on the dimensions Of interpersonal complementarity, content anal- ysis, and relationship perceptions in supervisory pairs. This chapter will examine (1) sample selection, (2) data collection, (3) instrumentation and (4) hypotheses tested. Sample Selection The subjects in this study were ten doctoral practicum trainees and their psychotherapy supervisors at the Michigan State University Counseling Center. The Counseling Center offers a variety of mental health services to the academic community. The services include individual and group psy- chotherapy, vocational counseling, and outreach consultation. The trainees were enrolled in both clinical psychology and counseling psychology Ph.D. programs. All student trainees had taken extensive course work and had progressed to at least the second year in their respective programs. Three of the trainees were enrolled in the clinical psychology pro— gram, with one having had a year of previous psychotherapy supervision. The other seven trainees were enrolled in the 40 41 counseling psychology program. With the exception of one term (ten weeks) of vocational pre-practicum supervision, these seven trainees had no previous psychotherapy super- vision in their doctoral programs. However all seven trainees had supervised experience in counseling work prior to entering their doctoral work. The ten supervisors were staff members at the University Counseling Center. Seven of them were senior staff members ranging in experience from approximately two to 25 years. Six of these seven supervisors held Ph.D. degrees in Clinical or Counseling Psychology, while the other senior staff member held a Master's degree in Social Work. The remaining three supervisors were intern staff members from doctoral programs in Counseling Psychology. Two of these supervisors had pre- vious intern experiences, while the other intern supervisor was serving as a first year intern. The interns had one year or less experience in the supervisor role. The pairing of supervisors and trainees was conducted by the training coordinators. Trainees were offered an Oppor- tunity to list preferred choices for a supervisor. However, with the trainee role also consisting of first and second year interns, practicum students were given last consideration in their choices. Therefore, pairing was primarily a random process. The supervisory pairs were established prior to their solicitation into this study. Ten separate supervisory pairs were selected without repetition of supervisors or super- visees to insure independence for statistical analysis. 42 Data Collection Four taped supervisor sessions were sampled from each pair over the course of the first two academic terms. Tapes were collected at the third week meeting (early first term), the eighth week meeting (late first term), the 13th week meeting (early second term) and finally on the 17th week meeting (late second term). The Barrett-Lennard Relation- ship Inventory was answered at the times of the first, third, and last taped sessions. With the interest of this study focused specifically on the early development of the supervisory relationship (begin- ning stage) through its middle working stage, it was assumed that the times selected for tape samples would be representa- tive Of the interactions in these two stages. The decision to sample these stages was supported by previous research indicating that the therapeutic relationship develops and progresses into its "working stage" over the first two stages (Mueller and Kell, 1972; Dietzel, 1971; McElhose, 1973). The theories of Mueller and Kell suggest that much of the work that occurs in therapy and supervision takes place in the middle stage of the process. Dietzel (1971) based his study on this expectation and found successful therapeutic dyads exhibit significant changes in levels of complementary interactions from the beginning stage of therapy to the middle stage. He inferred that lower levels of complementarity in the middle stage of therapy indicated greater amounts of therapeutic work had transpired. Mueller's (1969) research 43 revealed that at the beginning stage of therapy higher levels of complementarity would be evident. This was understood to be the result of maintaining interpersonal security within the newness of a relationship. Does (1969) and McElhose (1973) decided that the middle stage of supervision would best exemplify the supervisory work that they needed to sample for their studies. Therefore, it seems that if the develop- mental growth of a supervisory relationship will reveal itself in its interpersonal communications, it might be most evident in the first two stages. For the purpose of this study, each term (ten weeks) was arbitrarily identified as one Of the stages over the course of the academic year. The demarcation of stages by academic terms has been done in previous research on supervision (Does, 1969; McElhose, 1973). The personal experience of the writer in supervision and consultation with experienced supervisors in the academic setting lend support to the notion that most of the "work" done in supervision occurs by the end of the second term. However this will fluctuate from relationship to relationship and cannot be held as a rigid generalization. Nevertheless, due to time constraints and the availability Of sample selection it was necessary to arbitrarily establish the parameters within which this study would Operate. The type of "work" entered into by each supervisory pair was Of significance to this study. Capturing these work related interactions in the segments of each taped session was of primary importance. Therefore, 15 minutes of each 44 session (totalling one hour over four sessions) was taken for rating following the initiation of professionally related topics. Rather than aribrarily setting a time for sampling each session, the first 15 minutes related to "pro— fessional develOpment" was sampled. Those areas that fall under "professional develOpment" were defined as conversation evolving around: (1) client concerns, (2) personal concerns, (3) practicum concerns, and (4) academic concerns. The purpose Of this delineation was to insure that most of the interactions were not initial greetings or results of a sporting event. If the whole ses- sion was Of a superficial nature, the 15 minutes following the first ten minutes were taken. It would have then had significance for the study if a whole supervisory session was spent discussing issues irrelevant to professional de- velopment. In addressing the issue of tape sampling, Dietzel (1971) and McElhose (1973) found it best to sample after ten minutes had elapsed into each session, allowing time for initial greetings to occur. Sampling small segments of tape at early, middle, and late points in each session, McElhose (1973) found that most of the "work" occurred at the early point in each session. Taking McElhose's (1973) conclusion into con— sideration, a similar but modified method Of sampling was utilized for this study. Therefore, in the present study only one 15 minute segment was rated rather than three seg- ments per session as in the study by McElhose (1973). If a 45 specific session designated for sampling was cancelled or if a tape malfunctioned, the following session was substi— tuted in its place. This procedure circumvented difficulties that arose, and was used to maintain consistency in the selection process. Following the collection of all tapes, verbatim tran- scripts were prepared. These transcripts were then used in the rating process for two of the measures employed in this study. The Barrett-Lennard Relationship Inventory The administration of the Barrett-Lennard Relationship Inventory (see Appendix A) occurred after the first, third and last taped sessions. The Relationship Inventory has been utilized in various types of research in the area Of supervision (Santoro, 1970; Condon, 1978; Reddy, 1960; Desrosiers, 1967; Hansen and Barker, 1963; Karr and Geist, 1977). The 67 item questionnaire is based on the "necessary and sufficient conditions" for therapeutic change postulated by Rogers (1957). Having initial application to the investi- gation of therapy relationships, Barrett-Lennard (1959) noted, "The client's experience of his therapist's response is the primary locus of therapeutic influence in their rela- tionship. The preceptions result from the interaction of his own personality characteristics and attributes of the therapist's actual experience in relationship to him" (p.2). 46 In this study both supervisor and supervisee responded to the 67 items on the questionnaire. The responses were each individual's perception of how the other person demon— strated the four facilitative Conditions. The combination of their scores provided a measure of the perceived quality of their relationship. Responses to the Barrett-Lennard Relationship Inven- tory in supervision research have shown that initial ratings of supervisors by supervisees tend to be high, possibly indicating that trainees are "playing it safe" (Santoro, 1970). Therefore in this study the inStrument was admini- stered three times over the course of two terms. It was expected that the supervisees and supervisors would have more substantive support for their perceptions on the second and third rating because of their longer experience base. The presence of facilitative conditions have value in all relationships (Rogers 1957). The specific conditions the Barrett-Lennard Relationship Inventory attempts to measure are the following: Emphathic Understanding - The extent to which one per- son is conscious Of another person's immediate experience. Level of Regard - The affective aspect Of one person's response to another. Unconditionalipy of Regard - How little or how much variability there is in one person's affective response to another (the degree of constancy of regard). 47 Congruence - The degree to which one person is func- tionally integrated in the context of his relationship with another, such that there is absence of conflict and there is consistency between his awareness and his overt communi- cation (Barrett-Lennard 1959). The supervisor and supervisee described their percep- tions of the nature of the relationship by indicating the degree of agreement or disagreement with each statement. The provision of three grades of "yes" and three grades of "no" (+1, +2, +3, or -1, -2, -3) allow for variation in agreement or disagreement. Investigating the validity of this measure, Barrett- Lennard (1962) employed the services of five judges all of whom were client-centered counselors of varying levels of experience. The judges classified each item on the measure as either a positive or negative indicator of the variable in question, and gave a neutral rating to any item that they regarded as irrelevant or ambiguous. There was perfect agreement between judges at the level of claffifying an item positive or negative on all except four items. Three of these four items were eliminated. In addition to content validation, an analysis of each item was undertaken as well as a split half reliability assessment of the entire instru- ment. Barrett-Lennard found this instrument to be a valid and reliable measure of the underlying theoretical model. The Barrett-Lennard Relationship Inventory has had many applications in the field of research on supervision. Karr 48 and Geist (1977), Dbsrosiers (1968), and Santoro (1970) are a few of those researchers who have found this instru- ment a valid and reliable measure in their investigations of the supervisory relationship. Behavioral Analysis System One process of analyzing tapes in this study involved the interpersonal system of behavioral analysis deve10ped by Freedman, Leary, Ossorio, and Suczek (1955), Leary (1957), and LaForge (1963). This process of tape analysis has been implemented by various researchers (Crowder, 1970; Dietzel, 1971; MacKenzie, 1968; McElhose, 1973; Mueller, 1969; Mueller and Dilling, 1968; Schiller, 1977). The method of analyzing tapes required two tasks. The first task required scoring each response unit (an inter- rupted communication) of both supervisor and supervisee. The unit was then placed into one Of four quadrants defined by two orthogonally-positioned axes: a dominant-submissive axis and an affiliative-disaffiliative axis. Descriptive verbs for the four quadrants included: (1) dominate, teach, give, support (friendly-dominant); (2) love, cooperate, trust, admire, (friendly-submissive); (3) submit, condemn self, distrust, complain (hostile-submissive); and (4) hate, punish, reject, boast (hostile-dominant). (See Appendix B for fur- ther elaboration). The second task of the raters in identifying the appro- priate quadrant was to assess the emotional state created by 49 the participants in each interaction, because it was taken into consideration that individuals create an emotional state by the manner in which they deliver a remark. The mood that is created by the speaker contributes to the response of the listener. The rater's task was to empa- thize with the person who was receiving the communication and identify the "affective pull" in the communication (Freedman 1951). Interjudge Reliability During Training Two raters were trained to employ the behavioral analy- sis system by using tape segments of supervisory sessions that were not part of this study. These tapes consisted of interactions between supervisors and supervisees similar to those in the present research. Both raters were masters level therapists working in community mental health settings. One rater had eight years of individual, group, and family therapy experience, while the other rater had been working in similar professional activities for four years. As trained and experienced therapists both raters were sensitive to the subtle affective components accompanying interpersonal inter- actions. This was a necessary criteria for their selection, since raters must have adequate empathic ability to properly utilize the system of analysis. The training procedure involved rating supervisor and supervisee responses into the apprOpriate quadrants of the Interpersonal Circumplex depicted in Figure l (p.21). 50 Raters were provided practice sessions prior to taking actual reliability ratings. The sessions involved rating five minute tape samples and discussing interrater discrepan- cies in the behavioral analysis system. General interjudge reliability levels in past research have yielded reliability coefficients of .80 (Dietzel 1971), and .91 (McElhose 1973). Considering there was a .25 chance in agreement between raters (probability of identifying the same one of four quadrants), acceptable levels of reliability were expected. Raters simultaneously and independently rated identical responses on nine five minute segments of taped interaction. Each response item was gathered and percentages of agreement between raters was calculated. Researching client—therapist complementarity and role behavior, Schiller (1976) utilized a percentage of agreement as a measure of reliability between his raters employing the Leary Circumplex. He found that since the measure required double coding (i.e. one for each of the two axes of the cir- cumplex), the amount of disagreement between raters had to be calculated. This was best accomplished by calculating item by item agreement rather than overall agreement. Hollander (1979) employed a similar reliability computation for raters in his study of family interactions. The index of concordance as he referred to it, is virtually a percentage of agreement. Table 2 lists the percentages Of agreement between raters in the present study. 51 TABLE 3.2 Interjudge Percentage of Agreement on Training Sample Tape Response Items Number Percentage Segment Per Segment Disagreed of Agreement 1 23 3 87 2 27 5 81 3 20 l 95 4 10 0 100 5 20 l 95 6 9 l 89 7 9 O 100 8 ll 1 91 9 ll 1 91 140 13 Mean % = 91 Categorizing and Scoring Verbal Interaction The measuring device used for identifying and quantifying "therapeutic work" was the Hill Interaction Matrix (HIM), See Appendix C, p.6 . Developed by Copplino and Hill (1959), the matrix is used to conceptualize interactions along two dimensions: content (what is being discussed), and process (how it is being discussed). The HIM has been found to have application in various research on supervision and therapy (Lambert, 1974; Liebroder, 1962; Hansen, 1978). The matrix is composed of twenty cells, although only one quadrant (four cells) was used in the present study. 52 Lambert (1974) found this particular quadrant useful in the comparative investigation of supervisor therapeutic re- sponses in supervision and therapy. Descriptions of inter- personal exchanges that fall into the four cells are exemplified by the following: Relationship Confrontive - Members talk about the way they feel about each other and their relationship in a con— frontive way. e.g. "Jim, whenever I ask you a question you act like you're mad at me. Do I say things that rub you the wrong way?"‘ Relationship Speculative - In this cell members specu- late about how they view one another. A member may show that he wants to know how he is seen by another person. e.g. "Fred, are you feeling comfortalbe around me?" Personal Speculative - Here, a problem context of some kind has been established and peOple are working toward understanding of the topic person. e.g. "What are some of the situations where you run into this kind of problem?" Personal Confrontive - In this cell are statements that are an open, honest effort to get and give information. e.g. "You want your fiancee's time all to yourself; that's selfish." (Hill, 1960) The other 16 cells in the matrix involve less risk taking interactions, which range from "social amenities" type of conversation, to more defensive, "closing off" inter- actions. According to Hill when the interactions fall into 52 Lambert (1974) found this particular quadrant useful in the comparative investigation of supervisor therapeutic re- sponses in supervision and therapy. Descriptions of inter- personal exchanges that fall into the four cells are exemplified by the following: Relationship Confrontive - Members talk about the way they feel about each other and their relationship in a con- frontive way. e.g. "Jim, whenever I ask you a question you act like you're mad at me. Do I say things that rub you the wrong way?"‘ Relationship Speculative - In this cell members specu- late about how they view one another. A member may show that he wants to know how he is seen by another person. e.g. "Fred, are you feeling comfortalbe around me?" Personal Speculative - Here, a problem context of some kind has been established and people are working toward understanding of the topic person. e.g. "What are some of the situations where you run into this kind of problem?" Personal Confrontive - In this cell are statements that are an open, honest effort to get and give information. e.g. "You want your fiancee's time all to yourself; that's selfish." (Hill, 1960) The other 16 cells in the matrix involve less risk taking interactions, which range from "social amenities" type of conversation, to more defensive, "closing off" inter- actions. According to Hill when the interactions fall into 53 the four above listed cells, specifically when the process is speculative or confrontive and the content is personal or relationship oriented, the occurrence of "therapeutic work? may be assumed. The highest risk interactions are the relationship confrontive and personal confrontive ones. Rater Reliability on Hill Interaction Matrix The rating of interactions on the Hill Interaction Matrix was performed by a professional rater experienced in the use of this matrix. This person is the spouse of the developer and is active in the instrument's continued refinement and expanding application. She passed a criterion test following the completion of a course on HIM training (obtaining a perfect agreement score with the criteria), and has rated over a thousand therapy interactions mainly in connection with doctoral dissertations. The reliability in the present study was completed on 93 speeches randomly selected from the pool of supervisor/ supervisee interactions. Each speech had to be rated on both content style and work style dimensions in the matrix. Reliability estimates in previous studies employing the Hill Interaction Matrix were .92 for Lee and Bednar (1977) and .90 for Lambert (1974). Table 3 presents a breakdown of the percentage of agreement on the sampled interactions in this study. 54 TABLE 3.3 Rating Reliability Percentages on Sampled Interactions Total agreement on both content and work style 81 Agreement on content style but not work style 90 Agreement on work style but not content style 100 Mean reliability percentage for the present study 90 Hypotheses Hypothesis One: There will be a significant decrease in combined supervisory pair levels of com- plimentarity from the first stage of super- vision to the second stage. Mueller (1969) found that complementary interactions tend to be high at the beginning and late stages of therapy. Dietzel (1971) Observed that there was a significant de- crease in complementary interactions from beginning stage of therapy to the middle working stage in successful therapy dyads. Both of these studies and the theory underlying interpersonal complementarity as an attempt to maintain security and minimize anxiety, led this researcher to believe there would be a decrease in complementarity as relationships develop. Since the supervisory relationship is the unit of analysis in this study combined scores for supervisors and 55 supervisees were computed for the statistical analysis. A one-tailed t test was employed to statistically analyze differences in complementarity levels between stages I and II. Hypothesis Two: There will be a significant increase in combined supervisory pairs' therapeutic work responses (Hill Interaction Matrix, Quadrant Four) from the first stage (first term) of supervision to the second stage (second term). Kell and Mueller (1966) believe that the middle stage of therapy and supervision is a time for working through conflict. The early stage of supervision is spent fami- liarizing the supervisee with the process, and developing trust within the relationship. It was expected that there would be an increase in the amounts of higher risking interactions in the second stage when compared to the first stage. A one-tailed t test was employed to examine dif- ferences in the amounts of therapeutic work responses from the first stage to the second stage of supervision. Hypothesis Three: There will be no difference in combined supervisory pair perceptions of their re- lationship from the first stage of super- vision tO the second stage. 56 Alternative Hypothesis Three There will be a significant difference in combined supervisory pair perceptions of their relationship from the first stage of supervision to the second stage. It might be expected that as supervisory relation- ships mature the greater substance behind the mutual per— ceptions would lead to an increase in the perceived quality of the relationship. However, Santoro (1970) found that supervisees who rated their supervisors at the onset Of the relationship tended to rate high. This likelihood made it difficult to predict whether there would be any change in supervisee perceptions, or if changes in super- visor perceptions would counter the "halo effect" in supervisee ratings. Considering that other factors may also influence those perceptions, no directional hypotheses can be supported. Therefore a two-tailed t test was em- ployed to test for possible differences that may occur in supervisory pair combined perceptions from the first to the second stage of supervision. Hypothesis Four: There will be a significant negative re— lationship between combined levels of com- plementarity and combined therapeutic work responses in each Of the first two stages of supervision for supervisory pairs. 57 Dietzel (1971) indicated that in successful therapy dyads levels of complementarity decreased in the middle working stage when compared to the first stage. The in- ference made from his results revealed that therapeutic work occurred in greater amounts with successful therapy dyads who lowered their levels Of complementarity in the middle stage of therapy. This hypothesis was posited ex- pecting the possible inverse relationship between comple- mentarity and therapeutic work. For example, if a super- visory relationship generated high levels of complementarity in the communications between participants, indicating that anxiety was being avoided, it is expected that less anxiety producing therapeutic work responses would occur at that time. A Pearson Product—Moment correlation coefficient was calculated to investiage the possible inverse relation- ship between these two variables. Hypothesis Five: There will be a significant negative re- lationship between combined levels of complementarity and combined perceptions of the supervisory relationship in each of the first two stages of supervision for supervisory pairs. Complementarity in interpersonal communications can indicate that attempts are being made to minimize anxiety and maintain security within a given relationship. If that 58 particular relationship is perceived very positively on facilitative conditions, then those conditions may offer the support that is necessary to move into less complementary interactions. Rogers (1957) believes that these conditions are all that is "necessary and sufficient" for therapeutic change to occur. Dietzel (1971) found that when therapeutic change did occur complementarity was lowered. Therefore an inverse relationship was expected between these two variables. A Pearson Product-Moment correlation coefficient was calcu- lated to investigate this hypothesis. Hypothesis Six: There will be a significant positive rela- tionship between combined perceptions of the supervisory relationship and combined amounts of therapeutic work responses in each of the first two stages of supervision for supervisory pairs. Does (1969) investigated psychotherapy supervision on the dimensions of confrontation and facilitative conditions, and found in those relationships where conditions were high, more confrontation occurred. Anderson (1968) and Berenson (1968) discovered similar results with these two variables within therapeutic dyads. It was expected in this study that those relationships perceived highly on percep- tions of facilitative conditions, would engage in more higher risking therapeutic work responses. The statistical 59 analysis was implemented by employing a Pearson Product— Moment Correlation Coefficient to test this hypothesis. Analysis of Results As previously explicated under each hypothesis the following statistical analysis was employed in this study. Hypothesis one tested for decreases in complementarity from the first stage of supervision to the second stage. A one-tailed t test was utilized for that purpose. In the second hypothesis a one-tailed t test was used to test possible increases in combined therapeutic work responses for supervisory pairs from the first stage of supervision to the second stage. The third hypothesis was tested with a two-tailed t test examining possible changes in perceptions of relationship for supervisory pairs from the first to the second stage of supervision. In the final three hypotheses relationships between complementarity, therapeutic work, and perceptions of relationship were investigated in each of the first two stages of supervision. In all three of these hypotheses a Pearson Product-Moment Correlation Coefficient was employed to examine the relationships of all the variables in regard to each other. Results of Primary Concern The aforementioned hypotheses are meant to pertain specifically to supervisory pairs because of the perceived importance of the combined responses of the dyadal unit. 60 However, data will also be provided showing the results for supervisors only and supervisees only. The latter results will be utilized as additional informations to later explain possible discrepancies in findings for supervisory pairs and to provide notions about outher research possibili- ties. CHAPTER IV RESULTS Complementarity Hypotheses Complementarity was measured by identifying responses of both supervisor and supervisee according to one of the four quadrants of the Leary Circumplex. Once these re- sponses were categorized they were assigned their weighted value on the Shiller Scoring System. Averages (based on two taped sessions a term) were then computed for super— visors and supervisees, providing a complementarity index for each person. An overall index was then derived by com- bining the CI scores for each supervisoryypair. In addition, the same analysis computed for supervisory pairs has been computed for supervisors and supervisees. Although the analysis of data for the latter two groups does not relate directly to the research hypotheses of interest (supervisory pairs only), it was done in order to gain additional infor- mation and to help explain the findings of interest. Hypothesis I There will be a significant decrease in com- bined supervisory pairs' level of complementarity 61 62 from the first stage of supervision to the second stage. Table 4.1 presents the Complementarity Index (CI) scores for each supervisory pair for States I and II and then the amount and direction of change in these same scores. The highest possible CI score is 4.00. TABLE 4.1 Complementarity Index Scores* for Each Supervisory Pair for Stage I (Fall Term) and Stage II (Winter Term) STAGE 1 STAGE 11 CHANGE 1 3.54 3.57 + .03 2 3.88 3.60 - .28 3 2.52 3.35 + .83 g‘ 4 3.89 3.86 - .03 333 ->"4 5 3.86 3.89 + .03 MM 8ft 6 3.69 3.66 - .03 :1 m 7 3.98 3.76 - .22 8 3.70 3.59 - .11 9 3.68 3.66 - .02 10 3.56 3.66 + .10 *The higher scores indicate higher levels of complementarity. CI scores range from 2.52 to 3.98 for Stage I and from 3.35 to 3.89 in Stage II. Change scores range from -.28 to +.83. Only six of the ten supervisory pairs decreased in 63 complementarity as was hypothesized from Stage I to Stage II. Except for supervisory pair three the pairs had moderately high levels Of complementarity in Stages I and II. Table 4.2 shows the t values computed for the difference between CI mean scores for the combined group of supervisory pairs in Stage I versus Stage II. A one-tailed t test for correlated means was used for this purpose. The statistical analysis indicates that the mean CI scores for combined supervisory pairs did not differ significantly (t = -.28) between Stages I and II. Therefore the hypothesis that CI scores would decrease significantly for supervisory pairs from Stage I to Stage II was not supported. Table 4.2 also provides the added information concerning the mean CI scores and t values for all supervisors and all supervisees. Comparisons of mean CI scores between Stage I and II for supervisors and supervisees were statistically nonsignificant at the .05 level (t = -.17 and t = -.36 respectively for the two groups). Both supervisee and super- visor group CI scores increased from Stage I to Stage II. 64 TABLE 4.2 Complementarity Index Score Means, Change Scores and t Values for Supervisory Pairs, Supervisors, and Supervisees Across Two Stages Subjects Stage I Stage II Difference t Mean Mean Scores Values* Supervisory Pairs 3.63 3.66 + .03 - .28 Supervisors 3.64 3.66 + .02 - .17 Supervisees 3.62 3.65 + .03 - .36 *The t-value for 9 df that would be significant at the .05 level = 1.83 Therapeutic Work Hypothesis II There will be a significant increase in com- bined supervisory pairs' therapeutic work responses (Hill Interaction Matrix, Quadrant Four) from the first stage (first term) of supervision to the second stage (second term). Each independent response of supervisor and supervisee was assigned into its appropriate cell on the Hill Inter- action Matrix. Those responses categorized in the fourth quadrant of the matrix were the therapeutic work responses. The percentages of the total responses that were classified in this fourth quadrant were computed for each individual participant. Each Of the four tapings was calculated 65 separately. Next, the percentages of the two tapings within each term were averaged providing an average percentage for each stage for each individual. Finally, the percentages of each supervisory pair were averaged for the two stages. Hence, this computational process provided percentages of therapeutic work responses for each supervisory pair in the first and second stage of supervision. Inspection of the data in Table 4.3 provides evidence Of the amount of therapeutic work engaged in from the first to the second stage. Five supervisory pairs decreased their therapeutic work interactions in the second stage while the remaining five increased. It might also be noted that there is much variance across pairs in their percentages of thera— peutic interactions within each stage. The range in the first stage is from 47 percent to 90 percent, while in the second stage the range is from 38 percent to 94 percent. The range of percentage change from the first stage to the second stage is +13 to -26. A one-tailed t test for correlated means (see Table 4.4) was employed to statistically analyze differences in combined amounts of therapeutic work from the first to second stage of supervision for supervisory pairs. The t value (t = .45) was not significant at the .05 level. There was actually a slight decrease in the amounts of therapeutic work in the second stage. This was opposite to the expected direction. Therefore the hypothesis that therapeutic work would increase 66 significantly for supervisory pairs from Stage I to Stage II was not supported. A Comparison of Hill Interaction Matrix (Quadrant Four) TABLE 4.3 Mean Percentages and Difference Scores, for Individual Supervisory Pairs in Stages I and II Stage I Stage II Percentage (Mean Percentage) (Mean Percentage) Difference 1 90 94 +04 2 73 57 -16 3 64 38 -26 a 4 81 75 -06 §§ 5 52 65 +13 go. 6 69 72 +03 6 7 64 61 -03 8 83 90 +07 9 47 54 +07 10 87 86 ~01 The decrease in therapeutic work occurred also for both supervisors and supervisees when analyzed separately. visees as a group consistently entered into greater amounts of therapeutic work interactions in both stages when compared to their supervisors. Super- However the therapeutic work changes from Stage I to Stage II were not significant for either group at the .05 level. 67 TABLE 4.4 A Comparison of Hill Interaction Matrix (Quadrant Four) Mean Percentages, Difference Scores, and t-values for Combined Supervisory Pairs, Supervisors, and Supervisees Stage I Stage II Percen- t-value* (Mean (Mean tage Percen- Percen- Differ- tages) tages) ence Super- visory Pairs 71 69 -2 .45 Super- visors 66 65 -l .25 Super- visees 75 73 -2 .66 *The t-value for 9 df that would be significant at the .05 level = 1.83. Perceptions of Facilitative Conditions in Supervisory Relationships Mutual relationship perceptions were gathered from re- sponses Of supervisors and supervisees to the Barrett-Lennard Relationship Inventory. These individual responses were combined for each supervisory pair providing an index of the perceived quality Of their relationship. It was expected that changes would occur in these perceptions as the real relationships progressed over the first two stages of super- vision. 68 Hypothesis III There will be no difference in combined supervisory pair perceptions of their relationship (as measured by the Barrett- Lennard Relationship Inventory) from the first stage of supervision to the second stage. Alternative Hypothesis III There will be a significant difference in combined supervisory pair perceptions of their relationship from the first stage of supervision to the second stage. Table 4.5 presents data representing the combined scores Of supervisory pairs in each of the first two stages of super- vision. Seven Of the ten pairs show an increase in the per- ceptions of their relationship from the first to second stage. One remained the same, while the other two showed a decrease. Scores ranged from 3.5 to 4.9 for supervisory pairs in Stage I and from 3.6 to 5.3 in Stage II. Changes in Barrett- Lennard scores from the first to the second stage range from +6 to -6 for individual supervisory pairs. Table 4.6 shows the t values computed for the difference between relationship perception mean scores for the combined group of supervisory pairs in Stages I and II. A two-tailed t test for correlated means was used to test the null hypo- thesis. The statistical analysis indicates that the mean 69 relationship perception scores did not differ significantly (t = -.88) between Stages I and II. Therefore the null hypothesis stating that relationship perception scores would not differ significantly for supervisory pairs from Stage I to Stage II was supported. The data indicate however that there was a slight increase in the positiveness of relation- ship perceptions for supervisory pairs from Stage I to Stage II. TABLE 4.5 Comparison of Combined Relationship Perceptions of Individual Supervisory Pairs (as Measured by the Barrett-Lennard Relationship Inventory) for Stages I and II Stage I Stage II Difference Score 1 4.2 4.6 +.4 2 3.9 4.1 +.2 3 4.4 3.8 -.6 a t; 4 3.9 3.7 -.2 mcn Flu >44 5 4.9 5.1 +.2 u m aim % 6 4.7 5.3 +.6 m 7 4.7 4.9 +.2 8 3.5 3.6 +.l 9 4.3 4.3 0 10 4.4 4.5 +.1 70 Additionally, slight increases in the positiveness of relationship perceptions were noted for supervisors and supervisees from Stage I to Stage II. However these dif- ferences were not significant at the .05 level. TABLE 4.6 Comparison of Mean Relationship Perception (Measured by the Barrett-Lennard Relationship Inventory) Difference Scores, and t-Valued for Combined Supervisory Pairs, Supervisors, and Supervisees Stage I Stage II Percen- t- (Mean (Mean tage Value* Percen- Percen- Differ- tages) tages) ence Super- visory Pairs 4.29 4.39 +.10 —.88 Super- visors 4.10 4.22 +.12 -.86 Super— visees 4.48 4.56 +.07 -.75 *The t-value for 9 df that would be significant at the .05 level = 2.26. Relationship Between Complementarity and Therapeutic Work Hypothesis‘IV There will be a significant negative relation- ship between combined levels Of complemen- tarity and combined therapeutic work responses in each of the first two stages Of supervision for the supervisory pairs group. Additionally, 7O slight increases in the positiveness of relationship perceptions were noted for supervisors and supervisees from Stage I to Stage II. However these dif- ferences were not significant at the .05 level. TABLE 4.6 Comparison of Mean Relationship Perception (Measured by the Barrett—Lennard Relationship Inventory) Difference Scores, and t-Valued for Combined Supervisory Pairs, Supervisors, and Supervisees Stage I Stage II Percen- t- (Mean (Mean tage Value* Percen- Percen- Differ- tages) tages) ence Super- visory Pairs 4.29 4.39 +.10 —.88 Super- visors 4.10 4.22 +.12 -.86 Super- visees 4.48 4.56 +.07 -.75 *The t-value for 9 df that would be significant at the .05 level = 2.26. Relationship Between Complementarity and Therapeutic Work Hypothesis IV There will be a significant negative relation- ship between combined levels of complemen- tarity and combined therapeutic work responses in each of the first two stages of supervision for the supervisory pairs group. 71 Table 4.7 presents interrelatedness between comple- mentarity and therapeutic work responses for each of the two stages of supervision. Visual inspection of the data indicates that for both Stages I and II, complementarity scores tend to be quite high and the percentage Of thera- peutic work responses tend to remain moderate to high. SO in fact there tends to be little variance from Stage I to Stage II for each supervisory pair except for pair number three. The subjectively approximated line of best fit indicates a very slight positive relationship in Stage I, and a slightly stronger positive relationship in Stage II. In both stages this positive relationship between comple- mentarity and therapeutic work for supervisory pairs is opposite that which had been hypothesized. Additional tables are provided for the presentation of group data for supervisors and supervisees. Table 4.8 displays plottings of the relationship between levels of complementarity and therapeutic work for the supervisors group for Stages I and II. The plottings are mostly located in the upper right quadrant for both stages. Except for supervisor three all others had very high complementarity scores in Stage I. The complementarity score for supervisor three increased in line with the others in Stage II. In both Stages I and II the therapeutic work responses were moderate to high for supervisors. In Stage I the approxi- mated 1ine of best fit depicts a low negative relationship between scores on both variables. 71 Table 4.7 presents interrelatedness between comple- mentarity and therapeutic work responses for each of the two stages of supervision. Visual inspection of the data indicates that for both Stages I and II, complementarity scores tend to be quite high and the percentage of thera- peutic work responses tend to remain moderate to high. So in fact there tends to be little variance from Stage I to Stage II for each supervisory pair except for pair number three. The subjectively approximated line of best fit indicates a very slight positive relationship in Stage I, and a slightly stronger positive relationship in Stage II. In both stages this positive relationship between comple- mentarity and therapeutic work for supervisory pairs is opposite that which had been hypothesized. Additional tables are provided for the presentation of group data for supervisors and supervisees. Table 4.8 displays plottings of the relationship between levels of complementarity and therapeutic work for the supervisors group for Stages I and II. The plottings are mostly located in the upper right quadrant for both stages. Except for supervisor three all others had very high complementarity scores in Stage I. The complementarity score for supervisor three increased in line with the others in Stage II. In both Stages I and II the therapeutic work responses were moderate to high for supervisors. In Stage I the approxi- mated line of best fit depicts a low negative relationship between scores on both variables. 72 mmmnommg MACE 0.395 m0 005% 85850 __ m0<._.m - _ m0<._.m S5955 Boo 05 How DE ummm mo manna Egon; no mo gflgugfi pom HH pod H mmmmum 50m How manna Manson—am How. noncommmm NOB Oflsomfidfi Banal—8 can 5985838 no 393 858.8 6898 “1289.338 65. h.v a saxoos xapuI htmuanetdmo peurquoo -_ 73 noncommom xuoz Ofluoommnosa mo wmmucmouom . n. O.— n.— : Loin - .. 65m . mmzflmoonw puma on» How Dam umum No Down conned :xoummm am no aomumuumoHHH pom HH,Ocm H mommum nuom How muomflbuomsw How noncommmm xwoz Oflusummuwna can muwumucmEonEOU mo mHo>mA cmosuom mwcmcofiumHom one m.v. mqmflfi saloos xepux KarzeauaMeIdmoa 74 noncommom xuoz Owusommnone mo ommunmouom Z w0uom=m Hem mochOmum xuoz onusmmmnona can muaumucofioamaou mo mambuq cmmzumm magmGOHumamm use mne mamas 681003 xepuI Karlequamatdmoa 74 noncommom xhoz cansmmmuone mo ommucmouom __ m0uomom How noncommom xuoz Oeusommuose can mufinmacoEmamEOU mo mao>oq subsumm mesmGOflumamm one mfia mamas 881003 xepuI Karlequamatdmoa 75 In Stage II the subjectively approximated line of best fit indicates a low negative relationship between the two variables much the same as in Stage I. Visual inspection of Table 4.9 provides data on the relationship between levels of complementarity and thera- peutic work responses for supervisees for Stages I and II. The plottings for scores on these variables are located in the upper right quadrant in both stages. In Stage I six of the ten supervisees are gathered in the extreme upper right corner of the graph. This indicates high scores on both variables producing a ceiling effect. Supervisee number three is set off from the rest of the group and is located in the bottom left corner of the quadrant. The approximated line of best fit indicates a low positive relationship between the two variables in Stage I. A similar result is found in Stage II. Here the plottings are once again located in the upper right quadrant. How- ever, with the exception of supervisor number three the plottings tend to split Off into two groups. One group of plottings is located in the extreme upper right corner of the quadrant, indicating high scores on both variables. The other group of plottings is located at the top of the quad- rant but in a more moderately high range on the therapeutic work response variable. Supervisee number three is found to have moved to a plotting closer in proximity to the group when compared to Stage I, but still off by itself. The 76 approximated line of best fit indicates a moderate posi- tive relationship between levels of complementarity and therapeutic work for supervisees in Stage II. In both Stage I and II the positive relationships are Opposite to that which had been hypothesized. Table 4.10 provides the statistical analysis for the interrelationship between complementarity and therapeutic work within each stage of supervision. The analysis con- sisted of the computation Of a Pearson Product-Moment correlation for supervisory pairs, supervisors, and super- visees. For supervisory pairs, the correlation coefficients were not significant for either stage of supervision (r = .02 and r = .23, respectively). Thus the hypothesis that there would be a significantly negative relationship between com- bined levels of complementarity and combined therapeutic work responses in each of the supervision stages for the supervisory pairs group was not supported. The correlations between the two variables of interest for supervisors were r = -.25 and r = -.23 for Stages I and II respectively. Though these correlations are not signi- ficant they are negative and in the hypothesized direction. The supervisee group had correlations of r = .21 and r = .40 for Stages I and II respectively. These correlations were not significant and like the supervisory pairs group oppo- site tO that which had been expected. 77 TABLE 4.10 Pearson Product-Moment Correlation Coefficients* for the Relationship Between Complementarity and Therapeutic Work Responses for Supervisory Pairs, Supervisors, and Supervisees During each Stage of Supervision Stage I Stage II Supervisory Pairs .02 .23 Supervisors -.25 -.23 Supervisees .21 .40 *The Pearson Product—Moment Correlation for 9 df to be significant at the .05 level would be r = :.52 Relationship Between Complementarity and Supervisory RelatiOnship Perceptions Hypothesis V There will be a significantly negative rela- tionship between combined levels of comple- mentarity and combined perceptions of the supervisory relationship in each of the first two stages of supervision for the supervisory pairs group. Table 4.11 presents the plotting of the individual super- visory pair points representing the relationship between com- bined levels of complementarity and combined perceptions of supervisory relationships. This is done for both Stages I and II. Visual inspection of Stage I results indicate that, except for pair three, the group is located mostly in the 78 mesmcowumamm .HOmw>meom mo mGOADQOOHom pucwnfiou __ M0HOmm>Hmmsm How mflnmcoflumamm >HOmm>Hmmsm mo mCOAMQOOHmm pocmneou cam auflnmucmfimamfiou mo mam>mq pmcwneou :mm3umm menmcowumaom one HH.¢ mnm<9 saxoos xapux Karzequamatdmoa paurqmoo 79 mwsmcoHDMHum HHOmH>Mmmsm Ho m:OHumoouom __mouomsm How mHnmcoHumHom HHOmH>nwmsm Ho meowumooumm new muwumucoEOHmEoo HO mHm>uH cmmsuom mHnmcoHumHum one NH; ""393. 881035 xapuI Karlequamardmoo 80 mHsmGOHDMHom mHOmH>Hmmnm Ho maowummoumm mmcwmnowu mama How “Hm “mom HO OGHA poumEonummd am no :oHumuumoHHH cam HH cam H mommum zoom How nmomewmmsm How mHnmcoHumHom HHOmH>nmmsm Ho mcoHummouum mam muHHmucoEmHmfioo mo mH0>mH coo3uom mmnmcowumHmm one mHmv mAmHumsm mo chHummoumm pucHnEOO 86 axon thnadexaqm go abequaoxaa peutqmoa SBSUOOSBH : m0<._.m - m0Hom5m How mHzmcoHumHom muomflbummsw Ho maowumouuom OOGHAEOU pan noncommom xnoz OHusommumne pocHnEOO com3uom mHAmGOHDMHom one ma. .1 v mnmfla mflsmcowumHmm HHOmH>Hmmom mo mGOHummowum 87 sasuodsau xIOM otqnadexaqm JO abeauaoxaa . . mmaHmooum mama osu How uHm ummm mo OGHA pmemEonnmm4.:m mo :OHumHumsHHH . mam HH mam H mommum nuom How mHamH>Hmmsm How QHSmGOHumHmm HHOmH>ummsm mo mcoHummonmm pcm noncommmm xuoz OHummmmumna cmozuom mwnmcowumHmm one mane mflmfis aHamGOHumHom mHOmH>Huaam Ho maowuaoouoa 87 sasuodsaa 110M orqnadexeqm go abequaozaa . . mmaHaaonw mama may HOH uHa umma HO maHa pmumEonuaa¢.am Ho aOHumHumsHHH . cam HH tam H mummum zuoa How mHOmH>Hmasm How awnmaoHumHmm mHOmH>nmamm mo mcoHuaoouma mam mmmcoammm xuoB OHusmamumaa cumzuum aHnmcoHumHmm one m H. .. o mqmfifi. aHnmaoHumHmm >HOmH>Hmaam HO mQOHuamouua 87 sasuodsau XJOM otqnedezaum go afiequaozaa mmawaaouw mama may now uHa puma Ho maHa pmumEonnaa¢.am Ho GOHumHumaHHH . mam HH mam H mommum zuoa How mHOmH>Hwaam How awnmaoHumHmm >HOmw>Hmaam Ho mccHuaoouma mam mmmcoamom xuoz OHuamamHmne aoozumm aflcmaowumHoa one masw mandfi 88 aHsmaowumHum HHomH>Hmaam Ho maoHuauouma épé mmcfiasouw muma map How uwa umma HO DCHH pmumEonuaa< am Ho cOHumuumaHHH mam HH mam H mommuw zuoa now mommH>Hmasm How aHnmcowumHom anomH> Inoasm Ho mGOHuamouma mam mmmcoamom xHOB OHuamamumse subsuma aHamcoHumHmm one ha . v mqmdfi sasuodsaa xxom oggnedezaum go afiegueoxed 88 awsmaoHumHoa MHOmH>Hmaam Ho maowuamouua .0091, a no __moHoaaw now aHnmaowumHom HHOmH> inoasw HO chHuaoonma cam momcoamom xuoz OHusoamHoaa coozpma aHamcoHumHom one ha.v mamdfi sasuodsau axon oggnadegaqm go abegueoraa 89 various supervisors. Though the plottings continue to fall in or about the upper right quadrant, the approximate line of best fit shows a swing towards a moderately positive relationship between the variables. Supervisor number three shows the most marked change of any of the supervisors. In terms of therapeutic work responses for Stage II, super- visors one, seven, and nine stayed about the same as in Stage I. However, supervisors two, three, and four dropped, and supervisors five, six and eight increased. The relationship between combined percentage of thera- peutic work interactions and combined perceptions of super- visory relationship for supervisees is shown in Table 4.17. The plottings in Stages I and II can be seen to fall mostly into the upper right quadrant. The approximate line of best fit indicates a moderately high negative relationship be— tween the two variables in Stage I for the supervisee group. The results for the supervisee group in Stage II are similar but the negative relationship is more moderate. The vari- ability in the plottings for supervisees seems moderate in both stages but greater in Stage II. Table 4.18 shows the correlations betwen combined per— centage of therapeutic work interactions and combined per- ceptions of the supervisory relationship. For supervisory pairs, the correlation between the variables of interest for Stage I was r = -.53 and for Stage II r = .07. Hypo- thesis VI predicted a significantly positive relationship between therapeutic work and perceptions of the supervisory 89 various supervisors. Though the plottings continue to fall in or about the upper right quadrant, the approximate line of best fit shows a swing towards a moderately positive relationship between the variables. Supervisor number three shows the most marked change Of any of the supervisors. In terms of therapeutic work responses for Stage II, super- visors one, seven, and nine stayed about the same as in Stage I. However, supervisors two, three, and four dropped, and supervisors five, six and eight increased. The relationship between combined percentage of thera— peutic work interactions and combined perceptions of super— visory relationship for supervisees is shown in Table 4.17. The plottings in Stages I and II can be seen to fall mostly into the upper right quadrant. The approximate line of best fit indicates a moderately high negative relationship be— tween the two variables in Stage I for the supervisee group. The results for the supervisee group in Stage II are similar but the negative relationship is more moderate. The vari- ability in the plottings for supervisees seems moderate in both stages but greater in Stage II. Table 4.18 shows the correlations betwen combined per— centage of therapeutic work interactions and combined per- ceptions of the supervisory relationship. For supervisory pairs, the correlation between the variables of interest for Stage I was r = —.53 and for Stage II r = .07. Hypo- thesis VI predicted a significantly positive relationship between therapeutic work and perceptions of the supervisory 90 relationship in both stages of supervision. The results indicate that this hypothesis must be rejected. In fact, the correlation between the two variables was significant at the .05 level in the direction opposite that was pre- dicted for Stage I. Additional results for the supervisor group showed a moderate negative correlation (r = -.35) for Stage I and a moderate positive correlation (r = .41) for Stage II. Neither of these correlations was significant. For the supervisee group, the correlation between therapeutic work and perception of supervisory relationship (r = —.58) was significant at the .05 level. It was, however, in the direction opposite that predicted. The correlation for super— visees for Stage II was r = -.36 and was not significant. The results suggest a movement toward a less negative or more positive (however one wishes to look at it) relation- ship between therapeutic work and perception of supervisory relationships for supervisory pairs (r = -.52 to 4 = .07), supervisors (r = -.35 to 4 = .41) and supervisees (4 = -.58 to r = -.36) from Stage I to Stage II. 90 relationship in both stages of supervision. The results indicate that this hypothesis must be rejected. In fact, the correlation between the two variables was significant at the .05 level in the direction opposite that was pre— dicted for Stage I. Additional results for the supervisor group showed a moderate negative correlation (r = -.35) for Stage I and a moderate positive correlation (r = .41) for Stage II. Neither of these correlations was significant. For the supervisee group, the correlation between therapeutic work and perception of supervisory relationship (r = -.58) was significant at the .05 level. It was, however, in the direction opposite that predicted. The correlation for super- visees for Stage II was r = -.36 and was not significant. The results suggest a movement toward a less negative or more positive (however one wishes to look at it) relation- ship between therapeutic work and perception of supervisory relationships for supervisory pairs (r = -.52 to 4 = .07), supervisors (r = -.35 to 4 = .41) and supervisees (4 = —.58 to r = -.36) from Stage I to Stage II. 91 TABLE 4.18 Pearson Product-Moment Correlation Coefficients for the Relationship Between Combined Percentage of Therapeutic Work Interactions and Combined Perceptions of Supervisory Relationship Within Stages I and II for Supervisory Pairs, Supervisors, and Supervisees Groups Stage I Stage II Correlations Correlations Supervisory Pairs -.53* .07 Supervisors —.35 .41 Supervisees -.58* -.36 NOTE: A correlation of r = 1.52 would be necessary to achieve significance at the .05 level. Summary of Results Only one of the six hypotheses under investigation in this study was supported by the results for supervisory pairs. Hypothesis I which predicted a significant decrease in super- visory pair combined levels of complementarity from Stage I to Stage II was not supported. Hypothesis II predicted that combined percentages of therapeutic work interactions for supervisory pairs would increase significantly from Stage I to Stage II. The statistical findings failed to support this prediction. Hypothesis III stated that there would be no significant differences occurring in combined supervisory pair perceptions of relationship from Stage I to Stage II. The results supported the null hypothesis. In the remaining three hypotheses where relationships between the variables were investigated, the results were 91 TABLE 4.18 Pearson Product-Moment Correlation Coefficients for the Relationship Between Combined Percentage of Therapeutic Work Interactions and Combined Perceptions of Supervisory Relationship Within Stages I and II for Supervisory Pairs, Supervisors, and Supervisees Groups Stage I Stage II Correlations Correlations Supervisory Pairs -.53* .07 Supervisors -.35 .41 Supervisees -.58* -.36 NOTE: A correlation of r = :.52 would be necessary to achieve significance at the .05 level. Summary of Results Only one of the six hypotheses under investigation in this study was supported by the results for supervisory pairs. Hypothesis I which predicted a significant decrease in super- visory pair combined levels of complementarity from Stage I to Stage II was not supported. Hypothesis II predicted that combined percentages of therapeutic work interactions for supervisory pairs would increase significantly from Stage I to Stage II. The statistical findings failed to support this prediction. Hypothesis III stated that there would be no significant differences occurring in combined supervisory pair perceptions of relationship from Stage I to Stage II. The results supported the null hypothesis. In the remaining three hypotheses where relationships between the variables were investigated, the results were 92 not supportive of the directional hypotheses. Hypothesis IV which predicted a significant inverse relationship be- tween combined levels Of complementarity and combined percentages of therapeutic work responses for supervisory pairs within each stage was rejected. Hypothesis V which predicted a significant inverse relationship between super- visory pair combined levels of complementarity and combined perceptions of relationship within each stage was not sup- ported by the findings. Finally, Hypothesis VI predicted a significantly positive relationship between combined per— centages of therapeutic work responses and combined percep- tions Of the supervisory relationship for supervisory pairs in each stage. The results were not in support of the hypo- thesis. In fact, the correlation between the two variables of interest in Hypothesis VI was significant at the .05 level but in the direction opposite that predicted. Additional results paralleling those for the supervisory pairs group were calculated for supervisors and supervisees. In all cases except for Hypothesis VI, the statistical results for the supervisors and supervisees were not signifi— cant. For supervisees, like the supervisory pairs group, the correlation for the relationship between therapeutic work and perceptions of supervisory relationship was significant at the .05 level but in the direction Opposite that predicted by Hypothesis VI. CHAPTER V DISCUSSION The development of the psychotherapy supervisory rela- tionship on the dimensions of interpersonal complementarity, content of communications, and perceptions of facilitative conditions within the relationship was investigated. This investigation was conducted during the course of the first two stages of the supervisory process (the first and second terms of a three term academic year). Scores for supervisory pairs for levels of complementarity (complementarity indicies derived from the Leary Circumplex), therapeutic work re- sponses (Hill Interaction Matrix Quadrant Four), and percep- tions of facilitative conditions within the relationship (Barrett-Lennard Relationship Inventory) were examined. These variables were inspected for changes from Stage I to Stage II Of supervision and for their interrelatedness in each of the two stages. Hypothesis I addressed combined levels of complementary interactions for supervisory pairs in each of the first two stages of supervision. Complementary interactions are those where attempts to minimize anxiety and maintain interper- sonal security are evidenced in communications between 93 94 participants. A high complementary interaction would indicate a strong tendency for one individual to elicit a desired response from another individual and to have that response need met by the other. This first hypothesis predicted a significant decrease in the number of combined levels of complementary interactions from the first stage of supervision to the second stage. A one-tailed t test was employed to investigate the difference in means in each of the two stages for combined supervisory pair levels of complementarity. Results indicated no significant decrease from Stage I to Stage II. The complementarity interactions for supervisory pairs in both stages were relatively high, showing an actual slight increase in the second stage (see Table 4.2). A few major considerations are posited as explantions for this result. First, the expected changes in C.I. scores were investigated within the parameters of arbitrarily desig- nated time lines. These time intervals, Stages I and II, were established for the purposes of monitoring the "devel- Oping" supervisory relationship. The assumption underlying the hypothesis was contingent upon the temporal continuity of the developing relationship (i.e. that complementary interactions would decrease from the beginning stage of supervision to the middle "working" stage). The question arises whether these arbitrarily defined stages (ten week terms) truly represent a developmental time line equally applicable to each supervisory pair. Depending upon the 95 particular experiences of each supervisory pair, the "working stage" of their relationship may not have tran- spired until late in the winter term or possibly even spring. In another supervisory pair, the working stage may have eventuated much more quickly. Therefore, in testing this hypothesis which required sampling communica— tions representative of the first and second stages of supervision, academic terms may not have been true indicators Of the unfolding of those stages. Following this line Of thinking, each supervisory pair may develop in its own time sequence further defying the supervisory pair "group" pro- gression that was assumed in the design. A second point for consideration is whether or not signi- ficant changes in levels of complementarity which indicate stages in the therapy process (Dietzel, 1971; Mueller, 1969) can be equally applied to the supervisory process. Though there is similarity in the two processes, their progression may not develOp within similar time lines. In their investi— gations, Dietzel (1971) and Mueller (1969) found that levels of complementary interactions were a reliable measure of stages within the therapy process. Dietzel (1971) was able to predict success in therapy from low levels of complemen- tarity in the middle stages of therapy. Mueller (1968) detected high levels of complementarity at the early and late stages of therapy. An investigation of complementary inter- actions in supervision was undertaken by McElhose (1973). He sampled interactions during winter term and found no 96 significant changes in complementary levels within the course of each session. His findings indicated a lack of variance within each session similar to the lack of vari- ance in complementary interactions from Stage I to Stage II in the present study. It may be that the different nature of the work undertaken in therapy when compared to super- vision affects the time progression of the changing levels of complementary interactions. The client's willingness in psychotherapy to "let down" security operations so that they can be helped (thereby reducing the need to maintain complementarity) may be qualitatively different than the type Of communications exchanged in supervision. Whereas the client only risks self-disclosure and accompanying anxiety, all which may eventaully lead to the alleviation of pain, the supervisee may hold to more secure complementary interac- tions fearing reprisal and evaluation by the supervisor. Therefore, the supervisee might be more likely to maintain higher complementary interactions with the supervisor through- out the entire process than a client might maintain in ther- apy. The propensity to maintain higher levels of complemen- tarity in supervision may have caused or contributed to a ceiling effect in the scores. Referring back to Table 4.2 it is apparent that all scores for supervisory pair's, supervisor's and supervisee's groups ranged in the upper 12% of the scale (3.5-4.0 in a 4.0 scale) in both terms, al- lowing very little room for variance. This lack Of variance 97 coupled with the small sample size of the study further confounded the possibility of finding significance. Table 4 indicates that six of the ten supervisory pairs did de— crease in complementary scores during the second term. However, with pair three deviating far from the mean for the rest of the group in the first term (1.23 below the remaining group average), its eventual movement back into general conformity second term (increasing .89 to 3.3) Off- set the little variance there was in the changes of re- maining group scores. It should be noted though that even if pair three was eliminated, the little variance in scores for the other nine pairs and the limited power of a smaller sample would not have produced statistically significant results. The tendency for all scores to remain at the tOp of the complementarity scale deserves further comment. With interpersonal complementarity serving as a barometer for anxiety maintainance, the striving for what Lennard and Bernstein (1969) refer to as "homeostasis" in all relation- ships may lead complementarity levels to remain relatively high. A supervisory relationship in particular may be in- clined towards maintaining this balance because the super- visee is so invested in appearing as a competent professional. Therefore, there may be an optimal range of complementary interactions that score at the top Of the scale. This might indicate the effort Of both parties to primarily comply with one another, a necessary component for a good working relationship. The times of low complementary responses may 98 be the exception rather than the rule. Exceptionally high scores might indicate "walking on eggs" while an exception- ally low score may be a sign of unwillingness to establish rapport or incompatibility in the relationship. In the therapy relationship where Dietzel (1971) found the change from high to low scores in successful therapeutic dyads in the middle term, the change might have represented the move from an initial positive transference to a negative transference during the course Of therapy. The probability of a supervisee establishing a negative transference in supervision is less likely and not necessary or advantageous to the process. Also simply the notion that the supervisory relationship will last only so lopg may contribute to a lack of willingness to allow it to become more like therapy for both supervisor and supervisee. In Hypothesis II examination was given to the percentage of therapeutic work responses produced by supervisory pairs in the first and second stages of supervision. These thera- peutic work responses were interactions rated in the fourth quadrant of the Hill Interaction Matrix and were generally more interpersonally risking in nature. The hypothesis pre- dicted a significant increase in combined mean percentages of supervisory pair therapeutic work responses from the first stage of supervision to the second stage. A one-tailed t test revealed findings that showed no significant increase in the number of therapeutic work responses for supervisory groups from Stage I to Stage II. 98 be the exception rather than the rule. Exceptionally high scores might indicate "walking on eggs" while an exception- ally 1ow score may be a sign of unwillingness to establish rapport or incompatibility in the relationship. In the therapy relationship where Dietzel (1971) found the change from high to low scores in successful therapeutic dyads in the middle term, the change might have represented the move from an initial positive transference to a negative transference during the course of therapy. The probability of a supervisee establishing a negative transference in supervision is less likely and not necessary or advantageous to the process. Also simply the notion that the supervisory relationship will last only so lopg may contribute to a lack of willingness to allow it to become more like therapy for both supervisor and supervisee. In Hypothesis II examination was given to the percentage of therapeutic work responses produced by supervisory pairs in the first and second stages of supervision. These thera- peutic work responses were interactions rated in the fourth quadrant of the Hill Interaction Matrix and were generally more interpersonally risking in nature. The hypothesis pre- dicted a significant increase in combined mean percentages of supervisory pair therapeutic work responses from the first stage of supervision to the second stage. A one-tailed t test revealed findings that showed no significant increase in the number of therapeutic work responses for supervisory groups from Stage I to Stage II. 99 The expectation of finding greater amounts of thera- peutic work responses in the second stage of supervision is based upon the analogy of supervision to therapy again. Though the supervision process may at times resemble and take the form of therapy (Mueller and Kell, 1972), the quantity of therapeutic interactions seems to be much less. Furthermore, as previously explicated in the discussion of Hypothesis I, the progression of the supervisory relation- ship may evolve differently than that arbitrarily defined by the design of this study (i.e. stages = academic terms). Even if members of the supervisory pairs come to know each other better in time, the willingness to venture into higher risking interactions may not increase, but be main- tained at a constant level. The nature of supervision may incline participants to explore the intimacies of a client's life rather than to divulge much about themselves. Although experiential supervisors indicate a need for such occassional interaction, especially when the supervisee is involved in counter-transference webs with his client, in reality the confrontations may be kept to a relative minimum. This would support the research of Lambert (1974) who found that super— visors enter into significantly fewer therapeutic work re- sponses when doing supervision than when serving as therapists. When one considers the effect of extraneous pressures on the supervisee during winter term (Stage II), further light may be shed on the results under consideration. During that winter term, most practicum students at Michigan State 100 University are in their second year Of intensive coursework. They are also required to start pursuing internships for the following year, a process that requires much personal energy during that term. The combination of these and other taxing demands on the second year doctoral student, a year generally considered to be the most difficult, may lead the supervisee to seek safer ground and more support in supervision. With most of these demands occuring during winter term, the supervisee may want to utilize supervision as a place of refuge rather than one that creates more stress by venturing into higher risking interactions. The supervisor may well remember the situation in his own train- ing and collude with this possible unstated request. It is also likely that the supervisors are feeling high levels of stress at this time due to carrying heavy case loads, the depressive nature of winter, and the fact that many of the supervisors are asked to do a great deal of the training that takes place in the center. In short, by: the end of the second stage or winter term supervisors and supervisees may be experiencing "burn out". In addition to the stresses of the supervisee's second year as a confounding variable, a comment about the sampling of the interactions may be relevant to this discussion. The transcripts that were ultimately used for rating the HIM as well as the Complementarity Index were sampled near the beginning of the supervision sessions. Initially intending to follow a sampling design like that of Dietzel (1971) and 101 McElhose (1973) (i.e. sampling interactions at the begin- ning, middle and end points in a session), this researcher altered his design based upon a suggestion for future research made by the latter investigator. In his discussion of results, McElhose (1973) stated that most of the "work" in supervision seemed to occur at the early part of the session. The decision to sample the first 15 minutes of the supervisory session was based upon this observation. The question is now raised as to whether it might have been more advantageous to sample other points in the session, especially near the end when time to "warm up" may have transpired. Later interaction samples may have been more closely representative of the therapeutic work that tran- spired throughout the session and more sensitive to possible changes over the two terms. The analogy here is with the movement from Stage I to Stage II (working stage). The beginning parts of any session may be likened to Stage I and later parts of the supervisory session to Stage II. Also, using only one time sampling per session runs the same risk as having too small sample size--lack of repre— sentativeness, random findings, etc. Also, it is likely that each supervisory pair will move in and out of "work" phases within any one session which emphasizes the need for sampling more segments Of each ses- sion. Although the notion of time to warm up may have produced different rates of higher risking interactions within a 101 McElhose (1973) (i.e. sampling interactions at the begin— ning, middle and end points in a session), this researcher altered his design based upon a suggestion for future research made by the latter investigator. In his discussion of results, McElhose (1973) stated that most of the "work" in supervision seemed to occur at the early part of the session. The decision to sample the first 15 minutes of the supervisory session was based upon this observation. The question is now raised as to whether it might have been more advantageous to sample other points in the session, especially near the end when time to "warm up" may have transpired. Later interaction samples may have been more closely representative of the therapeutic work that tran- spired throughout the session and more sensitive to possible changes over the two terms. The analogy here is with the movement from Stage I to Stage II (working stage). The beginning parts of any session may be likened to Stage I and later parts of the supervisory session to Stage II. Also, using only one time sampling per session runs the same risk as having too small sample size--lack of repre— sentativeness, random findings, etc. Also, it is likely that each supervisory pair will move in and out of "work" phases within any one session which emphasizes the need for sampling more segments of each ses- sion. Although the notion of time to warm up may have produced different rates of higher risking interactions within a 102 session, it may not have equal application to the progres- sing relationship (warming up from first term to second term). Both supervisees and supervisors by the nature of their trade may have had a fair amount of experience dis- closing personally risky information. So to assume that the likelihood to move into more therapeutic work after two terms had elapsed in supervision may not be correct. The working stage in supervision may not be as distinctive in the middle stage as it is in therapy. The risk level a supervisor and supervisee choose to adOpt may be deter- mined much earlier in their relationship. This may then show little variance when inspected on a time line of two terms. Table 4.3 reveals that seven of the ten supervisory pairs varied seven percent or less in their total thera- peutic responses from Stage I to Stage II. At the same time however, the difference between high and low work response percentages among pairs was 47% for Stage I and 56% for Stage II. In summary there was a wide variation of therapeutic work responses among the supervisory pairs for either stage but most individual pairs remained relatively constant from Stage I to Stage II. It is interesting to note that of the interactions rated in the four cells comprising the fourth quadrant of the Hill Interaction Ma- trix, the majority of those interactions were in only two cells. These cells were the personal confrontive and per- sonal speculative. The other two cells which were the relationship confrontive and relationship speculative 103 scored less than 5% of all supervisory interactions. This might indicate that the therapeutic work responses occuring in supervision mainly concern one individual, probably most often the supervisee. The majority of pairs produced no relationship responses. This data may support another difference between supervisory and therapeutic relation- ships, even for those supervisors who endorse an experiential style of supervision. Furthermore, it may partially explain the results of Lambert's (1974) study where one practicing professional produced significantly fewer numbers of thera— peutic work responses while administering supervision than when doing therapy. If the situation noted by Mueller and Kell (1972) arises where the supervisee replicates in his interactions with his supervisor those same dynamics that his client exhibits with him in therapy, the situation is probably processed on an individual basis rather than by processing their own relationship as part of the resolution. In response to Hypothesis III, the facilitative condi- tions perceived by both individuals in the supervisory relationship were investigated. These conditions of empathy, congruence, positive regard and unconditionality of regard were measured by administering the Barrett-Lennard Relation- ship Inventory in each of the first two stages of supervision. The study examined possible changes in combined perceptions of the supervisory relationship from the first stage Of super- vision to the second stage. The results of this hypothesis indicated no significant change in either direction from 103 scored less than 5% of all supervisory interactions. This might indicate that the therapeutic work responses occuring in supervision mainly concern one individual, probably most often the supervisee. The majority of pairs produced no relationship responses. This data may support another difference between supervisory and therapeutic relation- ships, even for those supervisors who endorse an experiential style of supervision. Furthermore, it may partially explain the results of Lambert's (1974) study where one practicing professional produced significantly fewer numbers of thera— peutic work responses while administering supervision than when doing therapy. If the situation noted by Mueller and Kell (1972) arises where the supervisee replicates in his interactions with his supervisor those same dynamics that his client exhibits with him in therapy, the situation is probably processed on an individual basis rather than by processing their own relationship as part of the resolution. In response to Hypothesis III, the facilitative condi- tions perceived by both individuals in the supervisory relationship were investigated. These conditions of empathy, congruence, positive regard and unconditionality of regard were measured by administering the Barrett-Lennard Relation— ship Inventory in each of the first two stages of supervision. The study examined possible changes in combined perceptions of the supervisory relationship from the first stage of super- vision to the second stage. The results of this hypothesis indicated no significant change in either direction from 103 scored less than 5% of all supervisory interactions. This might indicate that the therapeutic work responses occuring in supervision mainly concern one individual, probably most Often the supervisee. The majority of pairs produced no relationship responses. This data may support another difference between supervisory and therapeutic relation- ships, even for those supervisors who endorse an experiential style of supervision. Furthermore, it may partially explain the results of Lambert's (1974) study where one practicing professional produced significantly fewer numbers of thera- peutic work responses while administering supervision than when doing therapy. If the situation noted by Mueller and Kell (1972) arises where the supervisee replicates in his interactions with his supervisor those same dynamics that his client exhibits with him in therapy, the situation is probably processed on an individual basis rather than by processing their own relationship as part of the resolution. In response to Hypothesis III, the facilitative condi- tions perceived by both individuals in the supervisory relationship were investigated. These conditions of empathy, congruence, positive regard and unconditionality of regard were measured by administering the Barrett-Lennard Relation— ship Inventory_in each of the first two stages Of supervision. The study examined possible changes in combined perceptions Of the supervisory relationship from the first stage of super- vision to the second stage. The results of this hypothesis indicated no significant change in either direction from 104 Stage I to Stage II. However, a slight positive increase was evidenced by seven of the ten supervisory pairs in their combined perceptions in Stage II. These findings are consistent with the research of San- toro (1970) who found that supervisees tended to rate their supervisors high at the onset of supervision. Although it might be expected that as time progressed later ratings would have greater substance and more experience to support them, it does not necessarily imply that these ratings would be more positive. As found by Santoro (1970), if the initial ratings by supervisees were high, the resulting "halo effect" would leave little room for reported increases over time. Overall trends did show that perceptions of both supervisor and supervisee groups (Table 4.6) increased from the first to the second stage. The changes in the ten supervisory pairs' perceptions in Tabel 4.5 indicate that seven pairs increased, one pair remained the same, and two decreased. Supervisors as a group tended to change their percep- tions nearly twice as much in the positive direction than did their counterparts. The majority of supervisors, having had more experiences in their roles than did supervisees in the trainee role, may have seen their trainees in a more realistic light at the outset of supervision. This factor would allow the supervisors more room for true variance (either up or down) in the following ratings. On the other hand, in addition to the superviSee's initially rating high to "play it safe" as Santoro (1970) suggests, they may have 105 had their perceptions clouded for other reasons. Many times a supervisee will "need" to see the supervisor as being more than he really is. This may be partly due to unconscious dependencies on the part of the supervisee, which distort his perception. As the real person behind the role of supervisor later becomes better known, those perceptions will become more realiStic. Whether they increase or decrease, the initial ratings may not indicate a true baseline for comparison. It seems that trying to independently investigate changes in perceptions within relationships over time involves many uncontrollable vari- ables. With each supervisor and supervisee rating being so intimately tied in a subjective reSponse to the other person, the many ups and downs in relationships may ulti- mately confound a true picture of perceptual changes over time. A rating of perceptions following a moderate period after the termination of the relationship might give a truer indication of how those perceptions altered over time. In regard to the Hypothesis IV the relationship between complementarity and therapeutic work responses was examined. It was hypothesized that there would be a significant nega— tive relationship between these two variables. As the need for maintaining high levels of complementarity decreased, the probability of more therapeutic work responses would in- crease. A Pearson Product-Moment Correlation coefficient revealed findings that were not significant. Actually a slight positive relationship was found between complementarity 106 and therapeutic work in both stages. These findings contra- dict the results found by McElhose (1973) where weighted relational distance (closeness between two participants in a dyad based on overt content areas discussed) increased when complementarity scores decreased. The findings also further distinguished style of communicating in supervision when compared to therapy. Dietzel's (1971) study found that greater success was achieved in therapy in those dyads where complementarity decreased in the middle working stage. The inference was made that when complementary scores de- creased more therapeutic work was taking place. In this study complimentarity scores and therapeutic work responses did not mutally exclude each other, but indicated that they both could be maintained at relatively high levels in a supervisory relationship. Once again the basic difference between the nature of supervision and therapy may account for the inconsistent results found in this study when compared to the Dietzel study. Table 4.10 indicates that the supervisor group did produce results showing a moderate negative relationship between complementarity and therapeutic work responses while the supervisee group responses vary differently. One possible explanation for these results may be that since the focus of supervision is generally upon the supervisee, his anxiety level is higher than the supervisor's, possibly motivating him to maintain more complementarity. Therefore, while both participants are engaging in personal speculative 107 or personal confrontive content categories on the Hill Interaction Matrix, it is probably the supervisee who is the object of the speculations and confrontations. His anxiety would be aroused, creating a situation in which a high level of complementarity would be desired by him. The supervisor may respond with equal complementarity but since it is not his work which is under scrutiny, he may engage in fewer therapeutic work responses. This dichotomy might explain the negative relationship between complemen- tarity and therapeutic work responses for the supervisor group when compared to the positive relationship exhibited by the supervisee group. It might be recalled that comple- mentarity levels for supervisors and supervisees were at similarly high levels in both stages (Table 4.2), while the supervisor group produced approximately 10% fewer thera- peutic work responses in each stage when compared to the supervisee group. The results of Hypothesis IV may imply that in super- vision, levels Of complementarity and percentages of thera- peutic work responses are not interdependent or mutually exclusive, but can operate independently of each other. It may be that the categorical dimensions of the Leary Circum— plex affect the likelihood of the therapeutic work responses occuring, rather than the level Of complementarity within those dimensions. For example, one will recall that the axes of the Leary Circumplex are divided into a friendly- hostile axis and a submissive-dominant axis. If one 108 supervisory pair were to engage in a predominantly hostile/ submissive (HS) to hostile/dominant (HD) interaction, and another pair engaged in a friendly/submissive (ES) to friendly/dominant (FD) interaction, both interactions would be scored highly complementary. However, one can imagine that the atmosphere of the friendly conversation would be much more conducive to risk taking than the hostile conver- sation. The probability of deducing the risk level in a conversation from the levels of complementarity exhibited in that conversation may be unachievable. Further, the therapeutic work captured by the Hill Interaction Matrix quadrant four may be qualitatively different than the type of "therapeutic work" inferred by Dietzel (1971). The Hill Interaction Matrix quadrant four measures personal risk in conversation while the therapeutic work Deitzel (1971) referred to may be much broader in sc0pe and include all the therapeutic skills employed by therapists. Hypothesis V focused on the relationship between com- bined perceptions of the supervisory relationship on its facilitative conditions and combined complementarity levels between supervisory pair participants. A significant in- verse relationship was expected between the two variables in each stage of supervision. The results did not support the hypothesis and actually indicated movement towards a more positive relationship between complementarity levels and perceptions of relationship. 109 By tracking the movement Of each variable independently from Stage I to Stage II, an explanation of the results of their correlational movement within each stage may be clari- fied. As previously noted, complementarity levels remained high in both stages. The responses to the Barrett-Lennard Relationship Inventory, while not changing significantly from Stage I to Stage II, did increase in seven of the ten pairs. This would account for the movement towards a more positive relationship between the two variables from the first to the second stage of supervision. In addition to explaining the results in terms of the quantitative changes of both variables in each stage, con- sideration of the theoretical assumptions underlying this hypothesis may Offer further insight. Rogers (1957) believed that the facilitative conditions measured by the Barrett- Lennard Relationship Inventory are all that is "necessary and sufficient" for change to occur in therapy. When Dietzel (1971) found therapeutic change to occur in therapy dyads, it was accompanied by low levels of complementarity. The under- lying rationale for Hypothesis V was based upon the expecta- tion that when facilitative conditions were perceived favor— ably the support experienced would lead supervisory pairs to lower levels of complementarity. However, a few assumptions made in the design of the present study may now clarify why the results did not indicate the expected inverse relation— ship. As previously mentioned the goal of supervision is not therapeutic change as it is in therapy, though some therapeutic 110 change may occur for the supervisee in the process. Also the nature of the relationship in supervision is different than that in thrapy. These differences confounded the investigatory jump made in the design of the present study. Very possibly too close an analogy between the two processes was made. Therefore, the results in the supervisory process were disimilar to those found in the therapy process. How- ever, the time constraints of supervision could be highly correlated with the inadequacy of the analogy. If there were no time limits to the supervision then the relationship and supervisee dynamics could be more fully explored leading to findings very similar to therapy. The assumptions made about complementarity deserve similar clarification. As mentioned in the discussion of Hypothesis IV the identified dimensions Of the Leary Circum— piex may have more bearing on the "climate" of a given inter- action than the complementary interactions along those dimensions. If a conversation is primarily along the hosti- lity axis Of the Circumplex it will probably have a different impact on the reported perceptions than a conversation adOpting a friendly tone. Therefore a friendly-dominant to a friendly-submissive interaction (typical of supervisory sessions in this study) may influence more positive percep- tions than a hostile-dominant to hostile-submissive inter- action. The confounding point is that both interactions are considered to be equal in their level of complementarity. 111 In Hypothesis VI the interrelationship between thera- peutic work responses and perceptions of relationship were examined. A positive relationship between these variables was predicted for supervisory pairs in both stages of super- vision. The findings revealed a significant negative rela- tionship between these two varibles for supervisory pairs in Stage I, and a slight positive relationship in Stage II. The supervisee group showed strong negative relationships between therapeutic work responses and perceptions of rela- tionship in both stages, significant in the first stage. The supervisor group evidenced a moderate negative relation- ship in Stage I, swinging to a moderate positive relationship in Stage II. These results were surprisingly contrary to the pre- diction. Client centered theory suggests that those relation— ships where facilitative conditions are the highest will produce the most therapeutic change. Although the initial research in this area focused on therapy relationships, Rogers (1960) suggested that the implications for supervision were equally important. If one looks at the supervisee and supervisor groups independently on these two dimensions possible explanations can be offered to clarify the results of this study. In the first stage, the supervisee group and supervisor group both showed a negative relationship between percen- tages of therapeutic work responses and perceptions of relationship. This might indicate that in the early stage of 112 their relationship both participant groups felt more posi- tively towards their counterparts when less risks were taken. The supervisees showed a significant negative rela- tionship between these two variables in this first stage. As previously discussed, the supervisory process is an evaluative process for the supervisee, and he may have seen his supervisor more favorably when he was not called upon to risk or self disclose. The supervisor may not have had personal evaluation weighing in the balance, but at the early phase of the relationship he too may have perceived the supervisee more favorably if he did not have to risk more in conversation. Stage I, however long it lasts, is likely to be a trust development stage and as such, less risking will be done. It may be that supervisors intuitively grasped this in the same way they would in therapy and wouldn't "overkill" with confrontiveness, in-depth interpretations, etc., unless a crisis situation existed. Stage I is where consideration of supervisee readiness for confrontation and correct timing of interventions takes precedence so that trust can be formed. Stage II shows a considerable turnabout for the super- visor group, while the supervisee group changed moderately. In the second stage the supervisor group revealed a moderate positive relationship between percentages of therapeutic work responses and perceptions of the supervisee. Super- visees still showed a negative relationship between the two variables, but it decreased in its strength. The following 113 may serve as an explanation for changes in the supervisor group. In Stage II the supervisors perceived the supervisees more positively when they increased their confrontation of the supervisees. They were taking more risks while they engaged in the difficult work of supervision. One might infer that because the supervisor perceived the supervisee more favorably on the facilitative conditions, he felt the security to trust that the supervisee would understand that his confrontations are a necessary part Of the process. A supervisee who is Open to growth and confrontation around his skills might be viewed more positively and invite the supervisor's initiating therapeutic work. The supervisor's perceptions of the supervisee may further increase as an aftereffect of a session where he felt responsible in his role to confront the supervisee. An example would be a difficult session where the supervisor may have had to address certain dynamics of the supervisee affecting his therapy and/or their supervision. The satisfaction following such a session or period of activity in their relationship may have eventuated in more positive perceptions of the supervisee. The less negative relationship between therapeutic work responses and perceptions of relationship for the supervisee group in Stage II may have resulted from a less defensive attitude adOpted over time. Though the supervisee still per- ceives the supervisor more favorable when not confronted or called upon to self disclose, his increasing positive 114 perceptions may indicate a growing trust in the supervisory process. Nevertheless, the supervisee's perceptions of the supervisor may still be motivated from a personal base which Hansen and Barker (1977) refer to as that which is "socially desirable" to the supervisee. That is, the super- visee may be conveying to the supervisor, "I perceive you more positively when you don't call upon me to take personal risks." Because of the extraneous pressures on the super- visee during training, he may honestly feel that the super- visor is more "empathic" to his situation by not creating more stress in supervision. The necessary confrontations and self disclosures may be received differently by supervisees than by clients. This may partly explain why differences are found in research on facilitative conditions and therapeutic change (or thera- peutic work responses, as is the case here) between the two processes. To reiterate a point made earlier, the motivation for self disclosure in therapy is different than in super- vision. The impetus to self disclose in therapy is usually from within. Despite its difficulty, the client in pain will Often feel better following such a release, and possibly reflect his gratitude with positive perceptions of the thera- pist. In supervision, the supervisee's need to self disclose may seem more externally demanded. The motivation in super- vision might be received by the supervisee as "another diffi— cult thing I mppp do to become an effective therapist." The perceptions of the supervisor following such an encounter 115 may initially suffer as a result. Also it is likely that supervisees wouldn't want the relationship to be too thera- peutic in nature because it might uncover what they fear would be flaws in their character which would threaten the pursuit of their profession. The phenomenal bias that ensues in such situations previously mentioned raises an issue of measurement. When drawing comparisons to the research of Does (1969), Mitchell, Monavac and Berenson (1968), and Anderson (1968) who all researched facilitative conditions in relationship to con- frontation, it must be kep in mind that the facilitative conditions in those studies were rated by outside Objective sources. It would have been interesting in this study to employ the Carkhuff (1967) scales to rate the facilitative conditions in addition to the participant's reported per- ceptions on the Barrett-Lennard Relationship Inventory. A measure of personal bias then could have been achieved. The bias might account for the different results in this study when compared to those found by Does (1969). In the latter study, supervisors who offered greater amounts of facilitative conditions were found to confront their super- visees more than those supervisors who rated low on the facilitative conditions. In the present research it might be inferred that those supervisors who confronted and en- couraged self disclosing interactions in their supervisees might have been rated lower as a result. This is 115 may initially suffer as a result. Also it is likely that supervisees wouldn't want the relationship to be too thera- peutic in nature because it might uncover what they fear would be flaws in their character which would threaten the pursuit of their profession. The phenomenal bias that ensues in such situations previously mentioned raises an issue of measurement. When drawing comparisons to the research of Does (1969), Mitchell, Monavac and Berenson (1968), and Anderson (1968) who all researched facilitative conditions in relationship to con- frontation, it must be kep in mind that the facilitative conditions in those studies were rated by outside objective sources. It would have been interesting in this study to employ the Carkhuff (1967) scales to rate the facilitative conditions in addition to the participant's reported per- ceptions on the Barrett-Lennard Relationship Inventory. A measure of personal bias then could have been achieved. The bias might account for the different results in this study when compared to those found by Does (1969). In the latter study, supervisors who offered greater amounts of facilitative conditions were found to confront their super- visees more than those supervisors who rated low on the facilitative conditions. In the present research it might be inferred that those supervisors who confronted and en- couraged self disclosing interactions in their supervisees might have been rated lower as a result. This is 116 understandable since one's perceptions of a supervisor might be unfavorably biased following a confrontation. These results might imply that when supervisors con- front their supervisees the supervisors will be perceived less favorably. Also, considering the much stronger nega- tive relationship between percentage of therapeutic work responses and perceptions of relationship in the first stage for both groups, some time elapsement may be advisable before serious encountering is initiated. One's sensitivity and defensiveness may need time to subside, allowing trust to develop. Therefore, it is possible that for most of the supervisory pairs, they were still, strictly speaking, in Stage I at the end of winter term. As in the case of most of the supervisory pairs in this study, if those relationships are perceived favorably over time, then more support will be experienced while engaging in the difficult interactions. This relationship would be consistent with the research of Berenson, Moravec and Mit- chell (1968) who found more self exploration following con- frontations in relationships where facilitative conditions were highest. Therefore, premature confrontations in a new relationship may create difficulty, while after a certain amount of trust has developed, they may be used constructively. Subsidiary Discussion Section It was apparent throughout the results of the study that while most of the supervisory pairs responded similarly, 117 supervisory pair three consistently moved in directions contrary to those of the rest of the group. Some specula- tion may offer clues as to why there may have been such a contrary response by this one pair. It is known that the supervisor was an intern with no previous supervisory experience (the least experienced supervisor in the group). As is Often the case, intern supervisors are not received with as much confidence by practicum supervisees when com- pared to their more experienced supervisor colleagues. More time may need to be spent gaining the supervisee's confi— dence. The other intern supervisors both had previous experience and revealed little deviation from the group. Therefore, this researcher feels the Obvious deviations of pair three warrant further elaboration. The following may offer evidence of a relationship that had initial difficul- ties, and hOpefully provide a measure of how those possible difficulties might be reflected in the variables of interest to this study. The changes exhibited by pair three on the various dimensions from Stage I to Stage II were found to be in Opposition to the first three hypotheses. It is interesting to note, however, that the interrelationships of the changes among the various dimensions for this pair did coincide with the predictions of the last three hypotheses. The results of the hypothesis investigating complementarity over time reveals the most obvious differences between this pair and the group as a whole. The average for all supervisory pairs 118 on levels of complementarity in the first stage was 3.6. With the exception of pair three, the group ranged from 3.5 to 4.0. The score for pair three on complementarity levels in the first stage was 2.5, a considerable deviation from the mean. Within the second stage the group ranged in closer proximity on complementarity levels. The group complementarity mean in this stage was also 3.6. Pair three changed their score in Stage II to a 3.3 increasing .8 when the majority of the other supervisory pairs remained largely the same. The deviation from group norms by pair three is evi- denced in the other areas under investigation. The results of the amount of therapeutic work produced over two stages revealed continued differences between pair three and the balance Of the supervisory pairs. The average amount of therapeutic work involvement in the first stage for all other supervisory pairs was 72%. Pair three was near that average with a reported 63% of therapeutic work responses. Within the second stage however, the average for the other nine supervisory pairs was 73%. Pair three decreased to 37% recording the most striking change in either direction for all pairs. The considerable decrease was influential in lowering the overall group mean in Stage II to below that of State I. Six of the ten pairs increased amounts of thera— peutic work in Stage II. The remaining four pairs decreased amounts (i.e. 13%, 7%, 3%, and 27% for pair three). 119 Finally, the changes in perceptions of the quality of supervisory relationships gave further indication of the distinctiveness of pair three, when compared to the rest of the group. Initial perceptions gathered at the onset of the supervision year for the other nine supervisory pairs was 4.3 on a 6.0 scale. Pair three was near this average with combined perceptions recorded at 4.4. Second stage scores for the group excluding pair three increased to an average of 4.5. Seven of the ten pairs increased the "posi- tiveness" in how they viewed the quality of their relation- ships. Of the remaining three, one pair decreased slightly by .2, the other remained the same, and pair three decreased considerably from 4.4 to 3.7 The interrelationships of scores on the three measures employed in this study of supervisory relationships were investigated in the last three hypotheses. Although pair three differed considerablyfrom the rest of the group in respect to its changes over time and the values of its scores, the interrelationship among the variables within its own movement was consistent with the last three hypotheses. When pair three's complementary levels increased from the strikingly low levels of the first stage to more typical levels in the second stage, its therapeutic work responses decreased by practically 50%. Although these changes were Opposite from the directions predicted, the inverse relation- ship initially hypothesized between complementarity and therapeutic work was evidenced. At the same time the combined 120 perceptions of their relationship fell Off considerably in the second stage when seven of the other nine pairs increased and one remained the same. A decrease in the perceptions of their relationship was accompanied by an increase in complementarity (consistent with hypothesis five), and with a decrease in therapeutic work (consistent with hypothesis six). Some insight may be gained by discussing this pair's movement throughout the study. With such low levels of complementarity in the first stage of supervision, pair three seemed to not have spent much time minimizing anxiety and establishing security Operations. Research on comple- mentarity has shown that high complementarity at the onset of relationships is typical (Mueller, 1968; Dietzel, 1971). It might even be considered appropriate as individuals es- tablish trust. The therapeutic work responses during Stage I for pair three was moderate. It may be assumed that risks were taken, but within an interpersonal style that offered little security. Pair three's Stage II scores might suggest a relation- ship that became problematic. There may not have been suffi- cient time spent interpersonally acclimating to each other in the early stage of their relationship. As a result it may be inferred that they retreated to safer ground in the second stage, as evidenced by considerably less risk taking and much higher levels of complementarity. Also, their 121 combined perceptions indicate a more negative perception of the quality of their relationship. The implications here may indicate that confrontations or high risk interactions taken too early in a supervisory relationship may be premature and lead to further compli- cations. Pair three may have exemplified the interrelation- ship Of such conditions. The high complementary scores of the other pairs may convey that a certain amount Of comple- mentarity is needed in establishing trust in a new super- visory relationship. As Mueller and Kell (1972) mentioned, a necessary amount of trust must develop for a supervisory relationship to mature. Suggestions for Future Research The design of this study was based upon a premise that drew an analogy between supervision and therapy. Though there is much overlap in the two processes, the present re- sults indicated that differences also exist. Future research might be directed towards uncovering these similarities and differences. This clarification may help the practitioner decide when to implement an experiential or didactic approach to supervision. The research might identify situations in supervision that would be more beneficially addressed in a therapeutic manner, employing an experiential approach. The tasks better served by a didactic approach could also be brought to light. Such an endeavor may provide a blueprint for appropriately deciding upon the separation or integration Of both supervisory approaches. 122 Further research on interpersonal complementarity may provide a better understanding of its processes. As a result of the present research, it was inferred that the categorical responses on the two axes of the Leary Circum- plex may have had as much influence upon the developing supervisory relationship as the dimension of complementarity along those axes. Further consideration could be given to the differences in complementarity on the hostility axis and on the friendly axis and how those responses interact with dominance and submission in relationships. In addition to a more detailed examination of comple- mentary interactions, various ways of measuring facilitative conditions in relationships might also be employed. For example, differences may be found in self reported percep- tions of facilitative conditions when compared to those objectively rated by outside observers. This would offer a measure of personal bias and identify the degrees of "halo effects" in ratings by supervisees of supervisors. It might also be interesting to note the degree of personal distortion on perceived conditions when compared to objective ratings and the implications they may have for identifying trans- ference in relationships. The perceptions of the quality of relationship may be given further support by a supervision satisfaction scale. The scale could be administered after an interum of time following termination. The time elapsement might provide for a period of integration, limiting the possibility of 123 situational or reactive bias in those ratings during the ongoing process. Also an interview with each participant during the ongoing process might monitor possible diffi- culties arising in the relationship not measured by the instruments. This would provide some understanding of the deviations exhibited by supervisory pair three. Future research should examine whether stages and school terms can be considered to be the same. If super- vision was open ended with no prescribed time lines, might it eventually come to approximate the therapy process? Had the present study incorporated spring term, the measures may have indicated a develOpmental progression similar to that which was hypothesized, but unfolding over a more extended period of time. Finally, mppe tape sampling (both within stages and sessions), subjects, and related measures will strengthen future studies on supervision. These steps will insure more representative supervisory interactions, capturing the con— structs of interest within those interactions, and providing more statistical power in the study. Summary The primary purpose of this study was to investigage the communications and perceptions of the developing super- visory dyad. Examination was given to the dimensions of interpersonal complementarity in communications between super- visors and supervisees, the amount of therapeutic work 124 interactions produced in those communications, and the per- ceptions of the facilitative conditions within the rela- tionship. These dimensions were investigated independently for possible changes from the first to the second stage of supervision, and in relationship to each other within each stage of supervision. The stages of the developing super- visory relationship were arbitrarily defined as the terms in a three term academic year. The subjects in this study were ten doctoral practicum students and their supervisors. All subjects were affiliated with the Michigan State University counseling center training program. Two taped sessions were taken four times, at the beginning and the end of the first and second terms of the academic year. An inventory was completed in the first and second term. From the tapes, ratings were made on two measures. The first was a measure for behavioral complimentarity. Ac- cording to Leary (1957) and Carson (1969), behavioral comple- mentarity refers to the degree of reward (i.e. interpersonal reinforcement) experienced by both participants in an inter- action as a result of the specific behaviors exhanged. High complementarity within a conversation indicates that anxiety is being maintain and interpersonal security is resulting. The second measure rated on the tapes was the Hill Interaction Matrix. The matrix is a system of categorizing responses on their content and process levels of communication. Of particular interest to this study was the fourth quadrant of 125 the matrix, where interactions indicated that therapeutic work was taking place. These were considered higher risking interactions, and they were noted to be personally or rela- tionship confrontive and speculative. The final instrument administered was the Barrett-Lennard Relationshiprnventory, The inventory measured perceptions of facilitative condi- tions reported by each participant. These conditions were noted by Rogers (1957) to be all that is necessary and suffi- cient for therapeutic change to occur in a relationship. Those conditions were empathy, positive regard, confruence, and unconditionality of regard. The first hypothesis predicted a significant decrease in combined complementary interactions from Stage I to Stage II. The postulate was not supported by the data. Hypothesis II predicted a significant increase in therapeutic work re- sponses from Stage I to Stage II. Contrary results were found; slight decreases were actually evident. In the third hypothesis, no significant changes were found to occur in combined perceptions fo the supervisory relatioship from the first to the second stage of supervision. The null hypothesis was supported. In response to the remaining three hypotheses, examina- tion was given to ther interrelationships among complemen- tarity, therapeutic work, and perceptions of relationship. Hypotheses IV and V predicted significant negative relation— ships between complementarity and therapeutic work, and complementarity and perceptions of relationship, in each 126 stage of supervision. Neither hypothesis produced signifi— cant results. The last Hypothesis (VI) predicted a signi- ficant positive relationship between combined percentages of therapeutic work responses and combined perceptions of facilitative conditions in the relationship in each stage of supervision. The results failed to support the hypothesis, and a significant negative relationship actually occurred in the first stage. In discussing the results of the study, the difference in the basic natures of supervision and therapy were con— sidered. The similarities between the two processes were pertinent to the present desgin. Consideration was given to the different effects of the categorical quality of communi- cations as well as their complementarity within those cate- gorical axes. The effects of the categorical responses were considered in their relationship to perceptions of relation- ship and willingness to take personal risks. Next, the impact of high risk interactions upon perceptions of relation- ship was discussed. Finally, the need for clarification between the supervisory and therapeutic processes was sug- gested for further research. In addition, it was suggested that more time be allotted for monitoring the relationship with more frequent sampling of interactions within sessions and stages. APPENDIX A BARRETT-LENNARD RELATIONSHIP INVENTORY Below are listed a variety of ways that one person could feel or behave in relation to another person. Please con- sider each statement with respect to whether you think it is true or not true in your present relationship with your super- visor/supervisee. Mark each statement in the left margin according to how strongly you feel it is true or not true. Please mark every one. Write in +1, +2, +3; or -1, -2, -3, to stand for the following answers: +1: I feel that it is probably true, or more true than untrue. +2: I feel it is true. +3: I strongly feel that it is true. I feel that it is probably untrue, or more untrue than true. I feel it is not true. I strongly feel that it is not true. 1. He/she respects me. 2. He/she tries to see things through my eyes. 3. He/she pretends that he/she likes me or understands me more than he/she really does. 4. His/her interest in me depends partly on what I am talking to him/her about. 5. He/she disapproves of me. 6. He/she understands my words but not the way I feel. 7. What he/she says to me never conflicts with what he/she thinks or feels. 8. He/she always responds to me with warmth and in- terest--or always with coldness and disinterest. 9. He/she is curious about "the way I tick", but not really interested in me as a person. 10. He/she is interested in knowing what my experiences mean to me. 11. He/she is distrubed whenever I talk about or ask about certain things. 12. His/her feeling toward me does not depend on how I am feeling towards him/her. l3. He/she likes seeing me. 14. He/she nearly always knows exactly what I mean. 15. I feel that he/she has unspoken feelings or concerns that are getting in the way of our relationship. 16. His/her attitude toward me depends partly on how I am feeling about myself. 17. He/she is indifferent to me. 18. At times he/she jumps to the conclusion that I feel more strongly or more concerned about something than I actually do. I I I O.) N l—‘ I I 127 128 He/she behaves just the way that he/she is, in our relationship. Sometimes he/she responds to me in a more positive and friendly way than he/she does at other times. He/she appreciates me. Sometimes he/she thinks that I feel a certain way, because he/she feels that way. I do not think that he/she hides anything from himself/herself that he/she feels with me. He/she likes me in some ways, dislikes me in others. He/she is friendly and warm toward me. He/she understands me. If I feel negatively toward him/her he/she responds negatively to me. He/she cares about me. His/her own attitudes toward some of the things I say or do stop him/her from really understanding me. He/she does not avoid anything that is important for our relationship. Whether I am expressing "good" feelings or "bad" ones seems to make no difference to how positively-- or how negatively--he/she feels toward me. He/she feels that I am dull and uninteresting. He/she understands what I say, from a detached, objective point of view. I feel that I can trust him/her to be honest with me. Sometimes he/she is warmly responsive to me, at other times cold or disapproving. He/she is interested in me. He/she appreciates what my experiences feel like to me. HE/she is secure and comfortable in our relationship. Depending on his/her mood, he/she sometimes responds to me with quite a lot more warmth and interest than he/she does at other times. He/she just tolerates me. He/she is playing a role with me. He/she is equally appreciateive--or equally unappre- ciative--of me, whatever I am telling him/her about myself. He/she does not really care what happens to me. He/she does not realize how strongly I feel about some of the things we discuss. There are times when I feel that his/her outward response is quite different from his/her inner reaction to me. His/her general feeling toward me varies considerably. He/she seems to really value me. He/she responds to me mechanically. I don't think that he/she is being honest with him- self/herself about the way he/she feels toward me. Whether I like or dislike myself makes no difference to the way he/she feels about me. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. NOTE: 63. 64. 65. 66. 67. 68. 69. 129 He/she dislikes me. I feel that he/she is being genuine with me. Sometimes he/she responds quite positively to me, at other times he/she seems indifferent. He/she is impatient with me. Sometimes he/she is not at all comfortable but we go on, outwardly ignoring it. He/she likes me better when I behave in some ways than he/she does when I behave in other ways. He/she feels deep affection for me. He/she usually understands all of what I say to him/her. He/she does not try to mislead me about his/her own thoughts or feelings. Whether I feel fine or feel awful makes no difference to how warmly and appreciatively--or how coldly and unappreciatively he/she feels toward me. He/she regards me as a disagreeable person. What he/she says gives a false impression of his/ her total reaction to me. I can be very critical of him/her or very apprecia- tive of him/her without it changing his/her feelings toward me. At times he/she feels contempt for me. When I do not say what I mean at all clearly he/she still understands me. He/she tries to avoid telling me anything that might upset me. His/her general feeling toward me (of liking, respect, dislike, trust, criticism, anger, etc.) reflects the way that I am feeling toward him/her. He/she tries to understand me from his/her own point of View. He/she can be deeply and fully aware of my most pain- ful feelings without being distressed or burdened by them himself/herself. Appendix was derived from G.T. Barrett-Lennard, 1962, Psychological Monographs, p. 34-36. APPENDIX B SCORING MANUAL FOR INTERPERSONAL BEHAVIOR RATING SYSTEM General Considerations The interpersonal Circumplex, as it will be used in the present study, consists of four categories or quadrants into which all interpersonal behaviors may be rated. The four quadrants are defined by two orthogonal axes; a vertical axis representing the dimension of dominance-submission, and a horizontal axis for the affiliative-disaffiliative (freindly- hostile) dimension. A behavior is judged into a specific category by making dichotomous decisions on both axes. In addition, descriptive terms and example statements, to be listed subsequently, are available for each category. In rating the responses, several problems arise. First, affect and content (i.e., words) may, or may not, be congruent. For example, consider the supervisee statement, "I like you." If this statement is genuine it would be rated friendly-sub- missive (love). If it came after an interpretation which the client did not want to deal with it would be rated hostile- submissive (complain). To minimize the above problems, the following rule was established: affect takes predence over content. Secondly: within a given unit (uninterreupted speech) one or more shifts in feelings (emotional tone) are possible. For example, the client may begin his/her speech with an openly hostile statement (hostile-dominant) and then shift during the same speech to a self-condemning statement (hostile-sub- missive). Where this occurs multiple scorings are required. 130 131 For the above example, the scoring would be as follows: SE: H-D . . . H-S. Where there are more than two shifts in the same unit, only the initial and terminal behaviors will be rated. The ad- vantage of this procedure is that it permits a separate analysis of supervisee (or supervisor) as (l) respondent to the preceding elicitations of the other party (here, the initial response in the sequence is used), and (2) elicitor (stimulus) of subsequent response in the other (her the ter- minal behavior is considered). Thirdly: in various cases, raters may use fifferent levels of interpretation. To avoid this, interpretations should not go beyond the immediate context. Descriptive Terms and Example Statements for Each Category The following abbreviations will be used: Supervisor = SR, Supervisee = SE. Friendly—dominant (FD) category - to dominate, teach, give, support. 1. Dominate (direct, command, diagnostic probe, independent behavior). SR or SE changes subject, begins new tOpic, asks information-gathering ques- tions, is dominating, bossy. 2. Teach (advise, give Opinion, inform). SR or SE gives Opinion, acts as authority on subject, in- structs. 3. 4. 132 Give (help, interpret beyond conscious feelings). Example: "If you feel uptight next week we could meet twice", or "Your relationship with your girl— friend appears to be similar to the one you had with your mother". Support (sympathize, reflect feelings, reassure, generalize conscious feelings, approve, nurture, therapeutic probe). As a general fule, reflec- ting feelings, generalizing feelings, therapeutic probes (when rated here) must come after a state- ment which contained that data that is reflected, generalized, etc. Support and reassurance does not have this limitation. Friendly-submissive (FS) category - to love cooperate, trust, admire. 1. 2. 3. 4. Love (affiliate, identify with). Examples: "I really like you", and "I feel close to you". Cooperate (confide, agree, collaborate). SE co- operates with SR, works on problem, answers ques- tions, elaborates on reflective statements, agrees with. Trust (depend, ask for help). Example: SE says, "This problem arose which I hope you will help me with..." Admire (ask opinion, praise). Example: SE says, "What should I do?" or "You're the best therapist in the Counseling Center". 133 Hostile-submissive (HS) category - to submit, condemn self, mistrust, complain. I. 4. Submit (defer, obey). A. Submission is more to avoid confron- tation than to accept validity of statement (sometimes follows an argu- ment). Also, when client expresses extreme help- lessness without belief that therapist can help. A mere "Yeah", or "I guess so" response when the therapist is attempting to elicit an elaboration or after the therapist has made a statement about something. Condemn self (depressed, withdrawn). SE says, "I feel worthless", or "I'm no goo". SR says, "If I were a good therapist, you wouldn't have those feelings". Distrust (suspicious, skeptical). A. SR or SE expresses skepticism about other person or his staements. A "What?" fol- lowing a very clear statement, or a "Maybe." Suspicious about feelings, motive, etc., of other party. Example: "I don't know if you feel that way about me or not". Complain (rebel, nag, sulk, passively resist). A. Includes defensive maneuvers, angry withdrawls. 134 into silence, resistance expressed in passive ways. B. Silences of 15 secons or more where the previous response would suggest that the person is feeling hurt or angry. Hostile-Dominant (HD) category - to hate, punish, reject, boast. 1. Hate (attack, disaffiliate). SE says, "Go to hell". 2. Punish (be sarcastic, threatening). SE says, "Peo- ple are going to keep bugging me until I kill my— self". 3. Reject A. C. (withholding, competing, accusing). SE or SR rejects (in hostile tone) the previous statement of the other. Example: "No, that's not so". SE and SR are arguing, competing, accusing openly. SE or SR refuses a previous directive. 4. Boast (narcissistic, self-stimulating, intellectual- izing). A. Boastful statements. Example: "I got the hisghest grade on that last exam". Wandering, free-associating, conversation in which the speaker provides his own stimula- tion. Usually includes "lists" of events from the past week, rambling statements, etc. 134 into silence, resistance expressed in passive ways. B. Silences of 15 secons or more where the previous response would suggest that the person is feeling hurt or angry. Hostile-Dominant (HD) category - to hate, punish, reject, boast. 1. Hate (attack, disaffiliate). SE says, "Go to hell". 2. Punish (be sarcastic, threatening). SE says, "Peo- ple are going to keep bugging me until I kill my- self". 3. Reject (withholding, competing, accusing). A. SE or SR rejects (in hostile tone) the previous statement of the other. Example: "No, that's not so". B. SE and SR are arguing, competing, accusing openly. C. SE or SR refuses a previous directive. 4. Boast (narcissistic, self—stimulating, intellectual- izing). A. Boastful statements. Example: "I got the hisghest grade on that last exam". B. Wandering, free-associating, conversation in which the speaker provides his own stimula- tion. Usually includes "lists" of events from the past week, rambling statements, etc. 135 C. SE or SR intellectualizes. Examples: SE says, "I haven't worked out my Oedipal con- flict yet", or SR says, "What is it that's troubling you?" NOTE: Appendix was obtained from McElhose (1970), p. 112-115. APPENDIX C HILL INTERACTION MATRIX Introduction Our task is to describe what counselors are talking about and how they talk about it in terms of this matrix. 80., we must consider two separate dimensions: Content - What is the statement about? Spyie - How does the speaker say it? This means that each time the counselor speaks we must make two separate judgements, as to 1) content category, and 2) style category. The Content Categories What is the statement about? The universe of what people talk about is divided into three broad categories. The titles for these categories are "Topic", "Personal", and "Relationship". These titles are only buzz-words; the definitions are below. For convenience we will label them I (Topic), III (Personal) and IV (Rela- tionship). The definitions of these categories follow: Content Category IV (Relationship) - talks about, refers to, demonstrates or alludes to the immediate here-and-now relationship between the counselor and the client. This may include mention of past contact between the two. If the content of the statement (or conversation) is about a relationship, but not the relationship between this counselor and client, then the statement is coded in Content Category I (Topic). 136 137 If the content of the statement is specifically and explicitly about the client's individual thoughts, feelings, attitudes about a relationship then the content is coded in Category III (Personal). Content Category III (Personal) - The statement is about the client's self...about the client's actions, prob- lems or personality, etc. Occasionally, a counselor's statement may be about the counselor's own actions, problems, personality, etc. When the counselor's emphasis to the client is on "you", the con- tent is in Category III (Personal--i.e. this particular per- son). Content Category I (TOpic) - The statement is about virtually anything which doesn't fall into the other two cate- gories. This category includes an infinite variety of things talked about, such as ideas, issues, topics which may be either trivial or prodound. The statement may be primarily about this client's feelings, thoughts and attitudes toward that other person). Discussion or description of an outside event belongs in Content Category I. A Category I statement may be made with very strong af- fect (i.e. in an emotionally charged manner) or very weak affect (i.e. emotionally uncharged), but what matters is ypap is being talked about. One way to choose among the content categories is to consider this: counseling involves self-understanding for the clinet. Talking about Category III or IV content 137 If the content of the statement is specifically and explicitly about the client's individual thoughts, feelings, attitudes about a relationship then the content is coded in Category III (Personal). Content Categpry III (Personal) — The statement is about the client's self...about the client's actions, prob- lems or personality, etc. Occasionally, a counselor's statement may be about the counselor's own actions, problems, personality, etc. When the counselor's emphasis to the client is on "you", the con- tent is in Category III (Personal--i.e. this particular per— son). Content Category I (Topic) - The statement is about virtually anything which doesn't fall into the other two cate- gories. This category includes an infinite variety of things talked about, such as ideas, issues, topics which may be either trivial or prodound. The statement may be primarily about this client's feelings, thoughts and attitudes toward that other person). Discussion or description of an outside event belongs in Content Category I. A Category I statement may be made with very strong af- fect (i.e. in an emotionally charged manner) or very weak affect (i.e. emotionally uncharged), but what matters is 222E is being talked about. One way to choose among the content categories is to consider this: counseling involves self-understanding for the clinet. Talking about Category III or IV content 137 If the content of the statement is specifically and explicitly about the client's individual thoughts, feelings, attitudes about a relationship then the content is coded in Category III (Personal). Content CategornyII (Personal) — The statement is about the client's self...about the client's actions, prob- lems or personality, etc. Occasionally, a counselor's statement may be about the counselor's own actions, problems, personality, etc. When the counselor's emphasis to the client is on "you", the con- tent is in Category III (Persona1--i.e. this particular per- son). Content Category I (Topic) - The statement is about virtually anything which doesn't fall into the other two cate- gories. This category includes an infinite variety of things talked about, such as ideas, issues, topics which may be either trivial or prodound. The statement may be primarily about this client's feelings, thoughts and attitudes toward that other person). Discussion or description of an outside event belongs in Content Category I. A Category I statement may be made with very strong af- fect (i.e. in an emotionally charged manner) or very weak affect (i.e. emotionally uncharged), but what matters is wpap is being talked about. One way to choose among the content categories is to consider this: counseling involves self-understanding for the clinet. Talking about Category III or IV content 138 approaches self—understanding directly. Talking about Category I content approaches self-understanding indirectly. The Style Categories The universe of pp! peOple express themselves is divided into four broad categories. The titles for these categories are "Conventional", "Assertive", "Speculative", and "Pene- trating/Confrontative". These titles are buzz words; the definitions are below. For convenience we will label them B (Conventional), C (Assertive), D (Speculative), E (Pene- trating/Confrontative). These are described below. The style of interaction is independent of content (I.E. of what is talked about--anything can be talked about in any style). Style Category B (Conventional) -- Statements that are in general socially appropriate anywhere. May be devoid of content. Pleasantries and amenities. Affect is often joking or casual. Four our purposes, typical counselor statements such as "mm-hmm", "yes", "I see", and other such will be con- sidered Category B. You must judge content by listening to what the client has said beforehand. Style Category C (Assertive) -- These statement are certainly insistent. The affect (emotional tone) may be angry or not, it may be matter-of-fact. In Category C, the counselor may be asserting. EG. "To me that means..." Sometimes the conversation has the flavor of an argument, or disagreement (eg. "But, I think that..."). 139 Sometimes it is simple agreement with the other person's assertions. Styie Category D (Speculative) -- Speculative, ques- tioning, intellectual, or controlled approach to pertinent issues or problems for the client. Eg. "I wonder if...", "Do you think that...?", "...is that so?" The statement is problem oriented and offers hunches, theories, speculations, possibilities. Style Category E (Penetrating/Confrontative) —- State- ment penetrates to the significant central aspects of the problem. Classifies, synthesizes, provides or offers some resolution to a topic. Confronts the client with other data about client's behavior (Cell 8) or, counselor's own thoughts or feelings about the client (Cell 12). May reflect what the client has said in a clear, concise way. May interpret. (It doesn't matter if the interpretation is wrong or right). One way to distinguish among these Style Categories is to ask, "Is the statement problem-oriented; does it offer help towards client self-understanding (whether directly or in- directly?" If it does, then the statement is Category D or E, if not, then the statement style is B or C. The HIM (Revised for Dyads) The matrix we are using combines the Style and Content Categories defined above. 140 Content I-Topic III Personal IV-Relation- ship egg B-Conventional l 5 9 C-Assertive 2 6 10 D-Speculative 3 7 ll E-Penetrating/ Confrontative 4 8 12 The cells are numbered only for convenience in labelling and later for computer scorring. The numbers are only tags and don't indicate proferences. In making your judgements it will often be helpful first to consider the broader distinctions in the Content and Style Categories. That is does the statement approach client self- understanding directly (III, IV) or indirectly (I), and is the statement problem- (or concern-) focused, (D,E) or not oriented to an issue at hand (B,C)? These distinctions are diagrammed below: Approach to Client Self-Understanding Indirect I Direct III IV Not Problem B l 5 9 Focused ............................................ . C 2 6 10 Problem D 3 7 ll Focused ____________________________________________ 141 The Rating Process NOTE : l. 2. You must listen carefully to every statement. You must judge the Content (I, III, IV) of each client statement so that you can rate the thera— pist's response in case it is ambiguous. The Sryie of the client's statement is not important for our purposes. Don't waste your energy judging that. Simplistically, you judge content on whether the statement is about Up (IV), Me (III) or grem, That, Him, Her, etc. (*). If you need to hear something again, speak up and we will play it back. This isn't a contest. When people get tired they don't listen. You have to listen carefully to do this well. If you are not listening, if you are tired or distracted, call for a break. I don't care how many breaks you need so long as you are listening carefully while the tape is playing. Caution -- the letters and numbers are only for identification pruposes, not to indicate qualita- tive differences. They are a kind of shorthand that will be useful for computer scoring. We are only describing. Cell 7 is not "better than" 3, etc. We are just labelling "apples" and "oranges". Appendix was obtained from William Fawcett Hill, 1965; Hill Interaction Matrix Scoring Manual. BIBLIOGRAPHY BIBLIOGRAPHY Anderson, S.C., "Effects of Confrontation by High-and-Low Functioning Therapists." Journal of Counseling Psych- ology, ii (5), 1968. Arbuckle, D.S., "Supervision: Learning Not Counseling." Journal of Counseling Psychology, ii, 90-94, 1965. Barrett-Lennard, G.T. "Dimensions of Therapist Response as Causal Factors in Therapeutic Change." Psychologi— cal Monographs, 1962, 1e, No. 43, Whole No. 562. Berenson, B.G., Mitchell, K.M., and Moravec, J.A., "Level of Therapist Functioning, Patient Depth of Self-Exploration, and Type of Confrontation." Journal of Counseling Psy- chology, 1968, ii (2), 136-149. Boyd, J.D., "Counselee Staffing for the Laboratory Counseling Practicum." Counselor Education and Supervision, 1979, 19’ 77-79. Carson, R.C. Interaction Concepts of Personality, Chicago: Aldine Publishing Company, 1969. Condon, S.R., "Therapeutic Conditions and Feedback as Vari- ables of Supervisory Relationships in Counselor Educa- tion." Dissertation Abstracts International, 1978, fig (9-A0, 5243} Crowder, J.E., "Transference, Transference Dissipation and Identification in Success Vs. Unsuccessful Psychotherapy." Unpublished doctoral dissertation, Michigan State Uni- versity, 1970. Carkuff, R.R. Helping and Human Relations, New York: Holt, Rinehart and Winston, 1969, 2 Volumes. Desrosiers, F.X. "A Study of the Personal Growth of Counseling Trainees as a Function of the Level of Therapeutic Con- ditions Offered by Their Supervisory Groups and by Their Supervisors." Dissertation Abstracts, 1968, 32 (l-A), 119-120. Dietzel, C.S. "Client-Therapist Complementarity and Thera- peutic Outcome." Unpublished doctoral dissertation, Michigan State University, 1971. Dittman, A.T., "Problems of Reliability in Observing and Coding Social Interaction." Journal of Consulting Psychology, 1958, 22, 430. 142 BIBLIOGRAPHY Anderson, S.C., "Effects of Confrontation by High-and-Low Functioning Therapists." Journal of Counseling Paych- ology, ii (5), 1968. Arbuckle, D.S., "Supervision: Learning Not Counseling." Journal of Counseling Psychology, ii, 90—94, 1965. Barrett-Lennard, G.T. "Dimensions of Therapist Response as Causal Factors in Therapeutic Change." Psychologi- cal Monographs, 1962, 12, No. 43, Whole No. 562. Berenson, B.G., Mitchell, K.M., and Moravec, J.A., "Level of Therapist Functioning, Patient Depth of Self-Exploration, and Type of Confrontation." Journal of Counseling Psy— chology, 1968, i2 (2), 136-149. Boyd, J.D., "Counselee Staffing for the Laboratory Counseling Practicum." Counselor Education and Supervision, 1979, i2, 77-79. Carson, R.C. Interaction Concepts of Personality, Chicago: Aldine Publishing Company, 1969. Condon, S.R., "Therapeutic Conditions and Feedback as Vari- ables of Supervisory Relationships in Counselor Educa- tion." Dissertation Abstracts International, 1978, 22 (9-A0, 5243. Crowder, J.E., "Transference, Transference Dissipation and Identification in Success Vs. Unsuccessful Psychotherapy." Unpublished doctoral dissertation, Michigan State Uni- versity, 1970. Carkuff, R.R. Helping and Human Relations, New York: Holt, Rinehart and Winston, 1969, 2 Volumes. Desrosiers, F.X. "A Study of the Personal Growth of Counseling Trainees as a Function of the Level of Therapeutic Con— ditions Offered by Their Supervisory Groups and by Their Supervisors." Dissertation Abstracts, 1968, ii (l-A), 119-120. Dietzel, C.S. "Client-Therapist Complementarity and Thera- peutic Outcome." Unpublished doctoral dissertation, Michigan State University, 1971. Dittman, A.T., "Problems of Reliability in Observing and Coding Social Interaction." Journal of Consulting Peychology, 1958, 22, 430. 142 143 Does, R.B. "A Process Analysis of Supervision in Psycho— therapy/Counseling Training: On the Dimensions and Levels of Supervisor Offered Conditions. Unpublished doctoral dissertation Michigan State University, 1969. Ekstein, R., and Wallerstein, R., The Teaching and Learning of Psychotherapy, New York: Basic Books, 1958. Freedman, Leary, Ossorio and Coffey, "The Interpersonal Dimension of Personality." Journal of Personality, Gray, K.K., "Supervision an Intellectual and Emotional Experi- ence." Paper presented at the American Educational Research Association, Minneapolis, Minnesota, 1974. Hansen, J.D., and Warner, R.W., Jr., "Review of Research on Practicum Supervision." Counselor Education and Super— vision, 1974, 10 (3), 261-272. Hill, William Fawcett. HIM Interaction Matrix and HIM: Inter- action Matrix Scoring Manual, University of Southern California: Youth Studies Center, 1975. Hollander, M., "Communication Patterns in Families with Acting Out, Depressed and Normal Adolescent Daughters." Un- published doctoral dissertation, Michigan State Univer- sity, 1979. Kell, B.L, and Mueller, W.J., Impact and Change: A Study_of Counseling Relationships, Appleton-Century-Crost, New York: 1966. Krumboltz, J. "Changing the Behavior of Behavior Changers." Counselor Education and Supervision, 1967, 6, 222-229. Kurtz, R., and Grummon, D., "Different Approaches to the Measurement of Therapist Empathy and Their Relationship to Therapy Outcomes." Journal of Consulting and Clinical geychology, 1972, 3, lO6-li6} Lambert, M.J., "Supervisory and Counseling Process: A Com- parative Study." Counselor Education and Supervision, 1974, i3, 54-60. Langs, R., The SupervisoryiExperience, New York: Jason Arson, 1979. Leary, T., Interpersonal Diagnosis of Personality. New York: Ronald Press, 1957. Lee and Bedner, "Effects of Risk Taking and Group Structure." Journal of Counseling Psychology, 1977, 24, 3, 191-199. 144 Lennard, H.L. and Bernstein, A., Patterns in Human Inter- action, San Francisco: Jossey Bass, 1969. Lister, J.L., "Counselor Experiencing: Its Implications for Supervision", Counselor Education and Supervision, 1966, 5, 55-61. Lower, R.B., "Countertransference Resistances in the Super- visory Situation, American Journal of Psychiatry, 1972, 129, 156-165. MacKenzie, H.M., "The Interpersonal Behavior of Normal and Clinic Family Members." Unpublished doctoral disser- tation, Michigan State University, 1968. McElhose, R.T., "Supervisor-Supervisee Complementarity and Relational Distance as Related to Supervisor Experience Level." Unpublished doctoral dissertation, Michigan State University, 1973. Miller, C.D. and Getting, E.R., "Students React to Super- vision." Counselor Education and Supervision, 1966, 1966, 6, 73-74. Morgan, S.M., "The Relationship Between Congruent Student and Supervisor Expectations, Student Performance, and Stu- dent Satisfaction in Counseling Practice." Disseration Abstracts International, 1976, November, Volume 32, (5-A) 2644-2645c. Mueller, W.J. and Kell, B.L., Coping with Conflict: Super- vising Counselors and Psychotherapists. New York: Appleton-Century-Croft, 1974. Mueller, W.J. and Dilling, C.A., "Studying Interpersonal Themes in Psychotherapy Research." Journal of Counseling PSXf chology, 1969, Volume 16, No. 1, 50—58. Patterson, C.H., "Supervising Students in Counseling." Journal of CounselingyPsychology, 1964, 11, 47-53. Pierce, R., Carkhuff, R.R., and Berenson, B.G., "The Differ- rential Effects of High and Low Functioning Counselors Upon Counselors in Training." Journal of Clinical Psye chology, 1967, g3, 212-215. Pierce, R. and Schauble, P., "Graduate School Training on the Facilitative Conditions: The Effects of Individual Supervisors." Unpublished Manuscript, 1969, Michigan State University. 145 Reddy, Brendan W., "The Qualitative Aspects of Feedback in Learning Empathy." Counselor Education and Supervision, g (1969), 176-181. Rice, T.K., "Supervision and the Counseling Practicum." Paper Presented at the American Educational Research Association, Minneapolis, Minnesota, 1974. Rogers, C.R., "The Necessary and Sufficient Conditions of Therapeutic Personality Change." Journal of Consulting PsycholOgY. 1957, 21, 95—103. Santoro, D.A., "Perceptions of Interview Behavior and Rela- tionships in Counseling Supervision." Paper Presented at the American Educational Research Association Con- vention, Minneapolis, Minnesota, March 2-6, 1970. Schiller, L.S., "The Intake Interview in Psychotherapy: Client-Therapist Complementarity and Role Behavior." Unpublished doctoral dissertation, Michigan State Uni— versity, 1977. Searles, H.F., "Problems of Psychoanalytic Supervision." In J.J. Masseurman (Ed.), Science and Psychoanalysis: Psyr choanalytic Education. Volume 5, New York: Grune and Stratton, 1962. Searles, H.F., "The Informational Value of the Supervisor's Emotional Experiences." Psychiatry, 1955, 1g, 135-146. Sullivan, H.S., Interpersonal Theory of Psychiatry. New York: W.W. Norton and Company, Inc., 1953. Sunbold, L.M. and Feinberg, B.L., "The Relationship of Inter- personal Attraction, Experience and Supervisors' Level of Functioning in Dyadic Counseling Supervision". Coun- selor Education and Sppervision, 1972, 11 (3), 187-193. Trembly, E.L., "An Outline of Counseling and Psychotherapy Supervision." Unpublished Manuscript, Central Michigan University, 1981. Truax, C.B. and Carkhuff, R.R., Toward Effective Counseling and Psychotherapy: Training and Practice, Chicago: Aldine, 1967. Walz, G.R. and Johnston, J.A., "Counselors Look at Themselves on Video Tape." Journal of Counseling Psychology, 1963, 10, 232-236. Wedeking, D.F., Scott, T.B., "A study of the Relationship Between Supervisor and Trainee Behaviors in Counseling 145 Reddy, Brendan W., "The Qualitative Aspects of Feedback in Learning Empathy." Counselor Education and Supervision, g (1969), 176-181. Rice, T.K., "Supervision and the Counseling Practicum." Paper Presented at the American Educational Research Association, Minneapolis, Minnesota, 1974. Rogers, C.R., "The Necessary and Sufficient Conditions of Therapeutic Personality Change." Journal of Consulting Psychology, 1957, 31' 95-103. Santoro, D.A., "Perceptions of Interview Behavior and Rela- tionships in Counseling Supervision." Paper Presented at the American Educational Research Association Con- vention, Minneapolis, Minnesota, March 2-6, 1970. Schiller, L.S., "The Intake Interview in Psychotherapy: Client-Therapist Complementarity and Role Behavior." Unpublished doctoral dissertation, Michigan State Uni- versity, 1977. Searles, H.F., "Problems of Psychoanalytic Supervision." In J.J. Masseurman (Ed.), Science and Psychoanalysis: Psy- choanalytic Education. Volume 5, New York: Grune and Stratton, 1962. Searles, H.F., "The Informational Value of the Supervisor's Emotional Experiences." Psychiatry, 1955, lg! 135-146. Sullivan, H.S., Interpersonal Theory of Psychiatry. New York: W.W. Norton and Company, Inc., 1953. Sunbold, L.M. and Feinberg, B.L., "The Relationship of Inter- personal Attraction, Experience and Supervisors' Level of Functioning in Dyadic Counseling Supervision". Coun— selor Education and Sspervision, 1972, 11 (3), 187-193. Trembly, E.L., "An Outline of Counseling and Psychotherapy Supervision." Unpublished Manuscript, Central Michigan University, 1981. Truax, C.B. and Carkhuff, R.R., Toward Effective Counseling and Psychotherapy: Training and Practice, Chicago: Aldine, 1967. Walz, G.R. and Johnston, J.A., "Counselors Look at Themselves on Video Tape." Journal of Counseling Psychology, 1963, 10, 232—236. Wedeking, D.F., Scott, T.B., "A study of the Relationship Between Supervisor and Trainee Behaviors in Counseling 146 Practicum." Counselor Education and Supervision, 15, 4, 259-265, June 1976. Winborn, B.B., Hinds, W.C., and Stewart, N.R., "Instructional Objectives for the Professional Preparation of Coun- selors." Counselor Education and Supervision, 1971, 10, 133-137. Wolberg, L.R., The Technique of Psychotherapy, New York: Grune and Stratton, 1954. “‘111111111111?